Discriminatory Clubs: The Geopolitics of International Organizations 9780691247809

Citation preview

d i s c r i m i n at o ry c lu b s

Discriminatory Clubs the geopolitics of i n t e r n at i o n a l o r g a n i z at i o n s

chr istina l . davis

princeton university press princeton & oxford

c 2023 by Princeton University Press Copyright  Princeton University Press is committed to the protection of copyright and the intellectual property our authors entrust to us. Copyright promotes the progress and integrity of knowledge. Thank you for supporting free speech and the global exchange of ideas by purchasing an authorized edition of this book. If you wish to reproduce or distribute any part of it in any form, please obtain permission. Requests for permission to reproduce material from this work should be sent to [email protected] Published by Princeton University Press 41 William Street, Princeton, New Jersey 08540 99 Banbury Road, Oxford OX2 6JX press.princeton.edu All Rights Reserved ISBN 978-0-691-24779-3 ISBN (pbk.) 978-0-691-24778-6 ISBN (e-book) 978-0-691-24780-9 British Library Cataloging-in-Publication Data is available Editorial: Bridget Flannery-McCoy and Alena Chekanov Production Editorial: Jenny Wolkowicki Cover design: Katie Osborne Production: Lauren Reese Publicity: William Pagdatoon Copyeditor: Bhisham Bherwani This book has been composed in Arno Pro Printed on acid-free paper. ∞ Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

For my parents

contents

Illustrations

xi

Acknowledgments 1

xv

Membership and International Cooperation

1

1.1

Defining IGO Membership

9

1.2

Membership in International Relations Theory

12

1.3

Geopolitical Alignment as Basis for IGO Cooperation

21

1.4 Chapter Overview

28

1.5

35

Conclusion

2 Flexibility by Design: Rules for Accession

3

37

2.1

IGO Accession as Club Membership

40

2.2

International Society and Ending IGO Membership

48

2.3

Data on IGO Accession Rules

53

2.3.1 Founding Charter Documents

54

2.3.2 Participation Mandate

56

2.3.3 Conditionality Terms

61

2.4 Toward a Broader Understanding of Accession

68

2.5

69

Conclusion

Membership Patterns in Economic Institutions

71

Christina L. Davis and Tyler Pratt 3.1

Geopolitics and Economic Cooperation

73

3.1.1

74

Testable Implications for Membership Patterns

vii

viii

3.2

3.3

contents

Empirical Analysis of IGO Membership Patterns

75

3.2.1 Data on Membership in Multilateral Economic Organizations

76

3.2.2 Logistic Regression Analysis of Membership

79

3.2.3 Finite Mixture Model of Weighted Decision-Making

85

Conclusion

89

4 Accession to the GATT/WTO

92

Christina L. Davis and Meredith Wilf 4.1 Supply and Demand of Membership 4.1.1 Discretionary Rules for GATT/WTO Accession

98

4.1.2 The Geopolitical Basis of the Multilateral Trade Regime

99

4.1.3 Examples of Accession Negotiations

101

4.2 Empirical Analysis of Entry into GATT/WTO

5

96

109

4.2.1 Data on GATT/WTO Application and Accession

109

4.2.2 Geopolitical Alignment and Trade Regime Membership

113

4.3 Conclusion

119

The OECD: More Than a Rich Country Club

123

5.1

Vague Rules and Selective Enlargement

127

5.2

Selecting for Similar Type in the OECD

129

5.2.1 The OECD Accession Process

129

5.2.2 The Price of Admission

132

5.2.3 Why Bother? Understanding Demand for Membership

133

5.2.4 Seeking Status through Association

136

Common Features of the Like-Minded Club

139

5.3

5.4 Statistical Analysis of OECD Accession

145

c o n t e n t s ix

5.5

Case Studies of OECD Accession

149

5.5.1

149

Mexico

5.5.2 Korea

152

5.5.3 Eastern Europe

158

5.6 Brazil as a Nonmember Partner

160

5.7 Conclusion

173

6 Japan’s Multilateral Diplomacy 6.1

Japan’s Membership in International Organizations

6.2 Entry into International Society

177 180 185

6.2.1 Communicating with the World: UPU and ITU

186

6.2.2 International Bureau of Weights and Measures

191

6.3 Acting like a Great Power: Japan in the League of Nations

194

6.3.1 Joining the League

194

6.3.2 The International Labour Organization

197

6.3.3 Exiting the League and ILO

199

6.4 The Return to International Society

202

6.4.1 GATT Entry

204

6.4.2 OECD Entry

209

6.5 Leadership in East Asia

214

6.5.1 The Tale of Two Banks: ADB and AIIB

214

6.5.2 From Follower to Leader: Japan in TPP

226

6.6 Japan and the International Whaling Commission

246

6.7 Conclusion

253

7 Club Politics in Regional Organizations 7.1

Defining Regions

257 259

7.2 Evolving Membership Patterns

265

7.3 EU: The Security Prerequisite for Entry

280

7.4 ASEAN: Noninterference Elevated to Security Principle

300

x

contents

7.5 The Missing Northeast Asian Regional Organization

309

7.6 Conclusion

310

8 Exclusion from Universal Organizations 8.1

313

The Political Logic of Exclusion

316

8.1.1

Gatekeeping over Sovereignty

320

8.1.2 Evaluating Conditions for Exclusion

327

8.2 Exclusion Mechanisms

328

8.2.1 Statistical Analysis of Accession to Universal Organizations 8.3 Korea

341

8.3.1 United Nations

341

8.4 Taiwan

347

8.4.1 World Health Organization

354

8.5 Palestine

360

8.5.1 UNESCO

364

8.6 Ostracism of South Africa under Apartheid 8.6.1 Expulsion from the UPU

368 373

8.7 Conclusion

379

9 Revisiting Anarchical Society 9.1

330

382

What We Have Learned

383

9.2 Geopolitical Discrimination and Cooperation

389

9.3 The Evolution of Cooperation

394

Bibliography Index 421

399

i l lu s t r at i o n s

Figures 2.1 Four models of IGO accession

43

2.2 IGOs by issue area

55

2.3 IGO size by issue area

57

2.4 Restrictions on participation mandate

60

2.5 Membership suspension and expulsion

65

2.6 Weighted voting

67

2.7 Accession patterns

68

3.1 Substantive effect of geopolitical alignment

82

3.2 Effect of alliances on IGOs in different issue areas

84

3.3 Geopolitical model estimates by year

89

3.4 Mixture model results for different samples

90

4.1 Membership in trade regime

93

4.2 Cost of delayed entry: China and Mexico

94

4.3 Comparison of UN ideal point distance

102

4.4 Variation in timing of accession to GATT/WTO

110

4.5 Substantive effect of UN voting similarity

116

4.6 Geopolitical alignment and accession rigor

120

5.1 Income comparison of OECD members and nonmembers

139

5.2 Democracy and OECD membership

140

5.3 Alliance ties of OECD members

141

5.4 UN voting and OECD membership

143

5.5 Estimated negotiation survival curves

146

xi

xii i l l u s t r at i o n s

5.6 Participation in OECD committees by nonmembers

161

5.7 Signature of OECD instruments by nonmembers

162

5.8 Signature of OECD instruments by Brazil

166

6.1 Japan’s IGO membership in comparison

180

6.2 Attention to ADB in Diet testimony

217

6.3 ADB Diet deliberation topics

218

6.4 Attention to AIIB in Diet testimony

223

6.5 Topics raised in AIIB Diet testimony

224

6.6 Views of China in AIIB Diet testimony

224

6.7 TPP partners in Diet testimony

242

6.8 RCEP partners in Diet testimony

245

6.9 Media attention to Japan and the IWC

252

7.1 Regional IGO charter restrictions on participation mandate

267

7.2 Estimates of membership probability in regional IGOs

272

7.3 UN voting comparison of the UK and Europe

298

8.1 Average membership rate in open eligibility IGOs

330

8.2 Substantive effect of geopolitical alignment on open IGO membership

337

8.3 Embassies and IGO membership

338

8.4 Diplomatic representation of North and South Korea

346

8.5 Taiwan’s membership in IGOs

351

8.6 Palestine income comparison

363

8.7 Geopolitical alignment and UNESCO vote for Palestine

366

8.8 Diplomatic ties of South Africa

369

8.9 South Africa’s membership in IGOs

371

Tables 3.1 Effect of alliances on IGO membership

80

3.2 IGO membership: Geopolitical vs. economic models

87

4.1 Estimates of time to GATT/WTO accession

115

i l l u s t r at i o n s xiii

5.1 OECD member enlargement

130

5.2 Effect of geopolitical alignment on OECD membership

148

6.1 Japan’s membership in IGOs

181

7.1 Effect of foreign policy on regional IGO membership

270

7.2 First differences of foreign policy on regional IGO membership

271

7.3 Analysis of regional IGOs by separate region samples (1)

274

7.4 Analysis of regional IGOs by separate region samples (2)

275

7.5 Regional IGO sample comparison

277

7.6 Voting and rivalry regression results

278

8.1 Effect of alliances on membership in open IGOs

334

8.2 Effect of alliances on membership with comparison of selection process

336

8.3 Embassy representation and IGO membership

339

8.4 South Africa’s membership in UN family IGOs

372

8.5 Taiwan’s membership in IGOs

381

acknowledgments

my first two books on international cooperation focused on the negotiation and enforcement of rules, but I was often asked about selection bias. That criticism led me to think more about how states choose their partners. This book attempts to provide the answer to why some states agree to pursue shared governance through joining international organizations. The story of Japan’s entry into the trade regime in 1955 revealed the importance of its alliance with the United States to open doors for a former enemy and mercantilist country in a decision that would lay the foundation for Japan’s export-led growth miracle. As I looked further, I found this was more than just an exceptional case. Across organizations and governments, relations between countries seemed to matter as much as or more than compliance with rules. While writing about accession to the WTO and OECD in papers that provided the impetus for my research on enlargement processes, the events in Ukraine during the fall of 2013 highlighted broader currents of meaning that attach to decisions about membership. For Ukraine, the choice between association with Europe in the EU Eastern Partnership Policy or deepening integration with Russia in a proposed Eurasian Economic Union sparked a social movement that overthrew the government and led to the sequence of events toward war with Russia. Clearly the participants understood that membership was about more than adjusting tariff rates. But I felt that theories on international institutions were neglecting this relational dynamic. Explaining how countries choose sides forces one to recognize the club dynamics that underlie multilateral organizations. Teaching international organization and global governance seminars at Princeton and Harvard has provided me with the opportunity to develop ideas in conversation with incredible students. As we debated the applications of Palestine or Taiwan to join international organizations or tried to understand Brexit, we discussed the many implications of membership for

xv

xvi a c k n o w l e d g m e n t s

the identity of states and their people. The students challenged conventional views about market failures and power politics and shared stories about their experiences in student clubs and sports teams. My office at Princeton where I did much of the research for this book overlooked the eating clubs on Prospect Street where students gather for meals and socializing, with each club having its own arcane system to select members. The “bicker” and “sign-in” processes for joining a Princeton eating club, which I heard about when I was doing research on negotiations over accession to the trade regime, offered the background for the analogy that runs through the book. Exchanges with students proved fertile ground to explore different approaches to international relations. My work with Meredith Wilf and Tyler Pratt began when they were students in my international organization Ph.D. seminar at Princeton, and led to co-authoring articles that greatly moved forward the research on this project. I could not have asked for better collaborators who brought creative ideas and statistical skills that made the research far better. I am grateful to them for letting me include their work as co-authored chapters 3 and 4, and I am deeply indebted to them for their input on the entire book. Policymakers in Japan, Korea, Europe, and the United States helped ground this research with real-world perspectives. I am grateful to the many policymakers who agreed to be interviewed and were generous to share their time and knowledge so that this book could better reflect the practice of diplomacy. Although some requested to remain anonymous and provided information on a background basis, I would like to thank in particular the following individuals: Byung-il Choi, Naoshi Hirose, Douglas Yu-Tien Hsu, Jun-il Kim, Takeshi Komoto, Frank Lee, Rokuichir¯o Michii, Seiichir¯o Noboru, Myoung-ho Shin, Noriyuki Shikata, Jonathan Sun, and Akihiko Tamura. For research in chapter 5, I benefited from access to the archives of the OECD, and I am grateful to the assistance of Sandra Willmott, who made it possible to gather extensive materials in a short visit in 2014. Several officials of the OECD were kind to grant their time for interviews and to share data on participation in the institution. I thank Gita Kothari, Luiz de Mello, Andreas Schaal, Jeff Schaffer, and Jan Schuijer. My research on the Taiwan case benefited from a visit to Taiwan in December 2019 to speak at the Taiwan Political Science Association and at National Taiwan University. I thank Hans Tung for providing me this opportunity and making comments on my research. Hans and May Lin of TECO Boston arranged interviews for me to meet with officials in government and the private sector.

a c k n o w l e d g m e n t s xvii

The chapters of this book were presented over the years at many conferences and seminars, where I received invaluable feedback with both sharp criticism and enough interest to encourage further work. I thank Inken von Borzyskowski, Lawrence Broz, Allison Carnegie, Jonathan Chu, Cecilia Corsini, Janina Dill, Axel Dreher, Marina Duque, Alison Duxbury, Andreas Fuchs, Judith Goldstein, Emilie Hafner-Burton, Marina Henke, Stephanie Hofmann, Yoram Hoftel, Liesbet Hooghe, Masahiro Kawai, Tsuyoshi Kawase, Keisuke Iida, Dan Keleman, Barbara Koremenos, Kyle Lascurettes, Phillip Lipscy, Tom Long, Ed Mansfield, Lisa Martin, Julia Morse, Megumi Naoi, Kalypso Nicolaïdis, Saadia Pekkanen, Jon Pevehouse, Mark Pollack, Paul Poast, Stephanie Rickard, Peter Rosendorff, Nita Rudra, Ken Scheve, Christina Schneider, Duncan Snidal, Mireya Solis, David Steinberg, Randall Stone, Mike Tomz, Felicity Vabulas, Lora Anne Viola, Erik Voeten, James Vreeland, Daniel Wajner, Oliver Westerwinter, and Boliang Zhu. I am surely leaving off some names, as this book has received countless rounds of feedback, and I am grateful to the helpful advice of all who have offered their questions and comments. Conversations with colleagues at Princeton and Harvard have enriched my understanding of international relations and offered a sounding board for different parts of this book. I am especially grateful to Stephen Chaudoin, Sannoy Das, Jeffrey Frieden, Amaney Jamal, Iain Johnston, Robert Keohane, Josh Kertzer, Sophie Meunier, Christoph Mikulaschek, Helen Milner, Andrew Moravcsik, Susan Pharr, Dustin Tingley, Mark Wu, and Keren YarhiMilo. Joanne Gowa pushed me with tough questions that were softened by friendship and a willingness to always give more comments. Robert Keohane encouraged me to probe more deeply into why relationships matter and to look for broader patterns that shape cooperation. I am grateful for institutional support from Princeton University for the starting years as I began research into the politics of membership and from Harvard University for the final years of the project. At Princeton, the Niehaus Center for Globalization and Governance provided rich intellectual exchange in the weekly seminars and in conferences around the world, and almost every chapter in this book was presented at these different venues. During the 2016–17 academic year, I hosted six scholars to visit Princeton for the Fung Global Fellows Program of the Princeton Institute for International and Regional Studies to research on international society and global governance. Lai-Ha Chan, Srividya Jandhyala, Anastassia V. Obydenkova, Jong Hee Park, Lena Rethel, and Vinícius Rodrigues Vieira contributed greatly

xviii a c k n o w l e d g m e n t s

to this manuscript through sharing their own research and providing me with feedback on the manuscript over the year. They joined Julia Gray, Anja Jetschke, and Krzysztof Pelc in a workshop for this manuscript in April 2017. The detailed comments on draft chapters at the book workshop offered important direction for finalizing the book, and I am so grateful for their careful reading and time spent reviewing each chapter. Julia Gray has been exceptionally kind to read multiple versions of many chapters over the years and exchange notes and reflections on everything from coding treaties to why all this matters. At Harvard, the associates and speakers of the Program on U.S.-Japan Relations have offered an insider perspective on key issues that have helped me connect my academic research to both historical events and current policy debates. Two years as a Radcliffe Fellow provided me with release from teaching and an incredibly diverse community. Learning different perspectives on discrimination as it has been experienced and studied across cultures and disciplines helped me sharpen my ideas on the tensions between universal norms and discriminatory practices that coexist within multilateral institutions. I am grateful to Susan and Kenneth Wallach for their generous funding to support scholarship at the Radcliffe Institute for Advanced Studies. For a book of this size, many research assistants deserve credit. Raymond Hicks provided extensive support for the development of the datasets analyzed in the book and spent countless hours delving into different sources to track down the charters of international organizations and any variable that could help improve my analysis. Max Calleo, Nina Coomes, Ashleigh Maciolek, and Emilee Martichenko worked as my faculty assistants to gather materials for case research and proofreading. Sophie Welsh handled everything from data coding to editing, as she learned more than she ever expected about international organizations. Many others have supported my research with critical input, and I am grateful to the following students: Idir Aitsahalia, Emmanuel Calivo, Brandon Chen, Mari Chen-Fiske, Ian Chong, Natalie Dabkowski, Nathan Eckstein, Cassandra Emmons, Noah Dasanaike, Tai Hirose, Armita Hosseini, Kaede Ishidate, Ali Jebari, Shir¯o Kuriwaki, James Lee, Jialu Li, Ruofan Ma, Kouta Ohyama, Alex Park, Yon Soo Park, Reshini Premaratne, Eli Soffer, Harry Sparks, Diana Stanescu, Shiyun Tang, Kento Yamada, Shun Yamaya, and Jason Weinreb. I thank the University of Chicago Press and the Southern Political Science Association for permission to include an edited version of “Joining the Club? Accession to the GATT/WTO,” which was published in the Journal

a c k n o w l e d g m e n t s xix

of Politics as a co-authored article with Meredith Wilf (2017, 79:964–978). I also thank Springer Press for allowing me to reprint part of “The Forces of Attraction: How Security Interests Shape Membership in Economic Institutions,” which was published in the Review of International Organizations as a co-authored article with Tyler Pratt (2021, 16:903–929). Portions of chapter 6 were previously published in “Japan: Interest Group Politics, Foreign Policy Linkages, and the TPP,” a chapter in Benedict Kingsbury et al. Megaregulation Contested: Global Economic Ordering After TPP, (Oxford University Press, 2019: p. 573–591) and appear by permission of Oxford University Press. At Princeton University Press, Bridget Flannery-McCoy guided the book from submission to completion, while Alena Chekanov and Jenny Wolkowicki helped with production and Heather Jones compiled the index. My copyeditor Bhisham Bherwani worked tirelessly to improve consistency and clarity throughout the manuscript. I am extremely grateful to the reviewers who saw promise in the ideas and urged clarification of key points to better explain my theory and contribution. Their detailed comments greatly improved the book. Trying to complete a book during a pandemic presented its own challenges, but I had already finished my field work, and I am fortunate to have a supportive family. My teenage children Keiji and Misaki have such a positive attitude about their own studies that I could not be discouraged by the various setbacks that came up in my work. I am grateful for their willingness to listen and share ideas over family meals and on our many hikes together. My parentsin-law Takashi and Fumiko Imai hosted us in Tokyo when travel was possible and made time for calls when we could not visit. My husband Kosuke Imai has supported me through every twist and turn of writing this book while raising a family. He both made sure I had time to write and forced me to stop writing so we could have fun together. Choosing Kosuke as my partner for life was the best decision I ever made—by any evaluation criterion. I dedicate this book to my parents, Alton and Carole Davis. They inspired my love of reading and learning and encouraged me to pursue my dreams even when it has taken me far away to explore the world.

d i s c r i m i n at o ry c lu b s

1 Membership and International Cooperation

member selection forms one of the defining processes of social organization. Whether it be individuals joining a sports team or a social club or countries joining international organizations, the choice of membership imposes categories on who we are and what we do. Group membership determines status. Theories of socialization and institutional commitment depend on our understanding the full context of rules. Even the fundamental difference between types of government derives from the selection process for leaders. Without understanding who takes part in governance, one cannot understand the effectiveness of the organizations that we study. Certainly we know that few organizations are formed around a random sample of actors— self-selecting, screening for talent, and selling access to the highest bidder represent some of the many ways through which groups form. How does member selection occur at the international level? The concern with selection bias has long confounded research on the effectiveness of international institutions because it is difficult to know whether the conditions leading to membership or the constraints of membership shape behavior. Few international organizations offer automatic admission. Even an organization like the United Nations that espouses universality encounters controversy over membership—North and South Korea were unable to join until 1991, Switzerland did not opt to join until 2002, and Palestine and Taiwan remain outsiders today. Nor do we see consistent enforcement of performance criteria with screening for quality. Indeed, some countries join organizations without making significant policy changes. How can we explain Turkey as a founding member of the Organisation for Economic Co-operation and Development 1

2

chapter 1

(OECD) in 1961 and communist Poland joining the General Agreement on Tariffs and Trade (GATT) in 1967? Even regional organizations constitute more than a geographic category, as seen by the British experience in joining the European Economic Community in 1973, 16 years after formation and its most recent decision to exit. Far from being automatic or technocratic, membership decisions are deeply political. States evaluate the benefits of joining the organization and their relationship with other members. Rather than treat biased entry into organizations as a nuisance for research on the effects of institutions, we must seek to understand the complex process by which states coalesce into groups that become members of an institution. This book develops a theory about international organizations as discriminatory clubs of states. A core of like-minded states with common security interests choose to cooperate on other issues through joining together in international organizations. Their geopolitical alignment shapes who wants to join an organization, whether they are accepted into the club, and the price of entry. Incontrasttotheoriesthatexplaincooperationintermsofmarketfailurewithin an issue area, my argument shows the channel by which security interests form the basis for cooperation and status. Geopolitical alignment generates the willingness to recognize the authority of another state as a rule-maker in global governance. Approving joint membership in international organizations allocates status. The empirical evidence highlights systematic biases in the patterns of states that enter international organizations. But this is not an argument about screening for compliance—states eschew the narrow selection criteria expected by a contractual approach in favor of discretionary selection as part of broad strategies of economic statecraft. Blackmail, side payments, and favoritism are rife in the accession politics of international organizations. Despite the emphasis on public goods as the core problem for cooperation, many policy problems present impure public goods for which clubs can provide the benefits to a limited number while excluding others. Most international organizations include provisions for defining membership as a way to restrict cooperation to a subset of states. Moreover, international organizations are discriminatory because members care not only about provision of the cooperative good but also about the identity of other members. Rather than screening for states with the highest capacity, the process of joining international organizations often resembles entry into a social club that admits friends and excludes rivals. Where social clubs may rely on race or socioeconomic status as the basis for discrimination, states use geopolitical alignment as the basis for discrimination in decisions over organizational membership.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 3

The discriminatory politics of membership embed within international society patterns of interaction. For example, not only NATO and other alliances, but also the European Union (EU), OECD, and international economic institutions like the GATT established boundaries of cooperation for a coalition of states that share a common security interest. The group of states commonly referred to as “the West” during the Cold War was defined by these overlapping memberships in institutions more than by a geographic location or liberal ideology. As a rising power, China pursues a strategy that first brings countries together under its umbrella for cooperation in institutions. From its co-leadership with Russia of the Shanghai Cooperation Organisation starting in 2002 to formation of the Asian Infrastructure Investment Bank in 2016, China unites like-minded states in institutions as a way to build a sphere of influence. Weaker states also use institutional agglomeration as a step toward unified action. The formal rules of membership draw lines to establish who is expected to act together, which is a critical component of social norms. The argument places security first in the sequence of cooperation. Yet the security interests that motivate coalition-building through shared memberships in institutions are more diffuse than specific threats to survival. Joining organizations is a tool of soft power diplomacy—organizational membership represents an investment in relationships with other states. In some cases, states offer preferential entry to a potential cooperation partner as a reward and bribe intended to entice them toward closer relations. Here institutions broaden and deepen the ties among states that are not allies. In other cases, shared membership helps states to consolidate alliance ties through expanding the range of issues for cooperation beyond security. Finally, rivalry manifests itself through excluding enemies from organizations as a strategy to deny access to both material benefits and relational networks. This sequential enlargement privileges early entry by allies and gradual enlargement to include others after the rules and membership coalition of allied states have already been consolidated. When allowed to enter, nonallied states may pay a higher price through larger reform commitments relative to the easy path to entry given to allies. Yet overexpansion may arise if too many join. The need to reconfigure which groups of states are willing to work together has become a factor driving the fragmentation of global governance across multiple overlapping international organizations. The theory looks beyond great power politics. Major powers are at the forefront of establishing institutions and serving as gatekeepers over membership, but they are not the only ones to use this strategy. Coalitions form through a

4 chapter 1

joint decision by states to form a group, and security interests shape not only the motives of powerful states who lead but also the willingness of others to participate. If anything, smaller states have greater need to use institutions as a way to expand ties and establish their position within the social hierarchy of the international system. For marginal states seeking to normalize relations after a rupture, entry into organizations signals commitment to joint actions and mutual recognition. The question of whether to join or leave a major international organization goes to the highest levels of authority within a state—these are not decisions left to mid-level diplomats. The Brexit referendum on UK membership in the EU absorbed national politics for years. But even for less high-profile cases, the decisions are taken very seriously. The Mexican debate over joining the GATT offers an interesting example. Having declined to join the original GATT agreement in 1948—the Finance Minister denounced it as a threat to national industries—the government began talks about joining in 1979. Working party negotiations were going well, with GATT members agreeing to allow flexibility for Mexico to preserve many of its core developmental state policies. Alongside those negotiations, Mexican president José López Portillo conducted a domestic evaluation on the merits of GATT accession, with input from economic analysts and in consultation with industry and labor groups. The question was covered in front-page media commentary (Ortiz Mena, 2005, pp. 221–222). Critics portrayed accession as a move toward dependence on the United States as well as a threat to economic sovereignty and the development model writ large. The final vote of the cabinet in March 1980 opposed accession. López Portillo announced the postponement of GATT accession on the forty-second anniversary of Mexico’s expropriation of U.S. oil companies, framing rejection of the GATT in terms of Mexican nationalism and anti-American policy (Story, 1982, p. 775).1 Eventually the government did join the GATT in 1986, after relations with the United States had significantly improved and economic reform held new urgency (Davis and Wilf, 2015). The foreign policy process at the domestic level reinforces the security framing of membership decisions. The executive leadership and foreign 1. Story (1982) contends that the desire to show independence from the United States perversely pushed López Portillo to decline joining after receiving strong pressure from the United States that it should become a member. López Portillo’s relations with the Carter administration were poor at this time.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 5

ministry prioritize security, and they are the lead actors who set the agenda to seek membership. Later, during ratification, other domestic actors must accept or decline membership as a package deal. Broad foreign policy identity augments the attractiveness of an organization that confers status by means of deepening association with other like-minded states. Or, as in the example of Mexico in the GATT, foreign policy tensions could reduce the appeal of joining. Actors who may have little understanding or interest in the details of the rules will pay attention to the high politics of joining an organization. Outside the scope of specific policy reforms and institutional constraints, membership plays a role in defining how states fit within international society. When protests erupted in Ukraine in 2013 over the choice between joining a customs union with Russia or signing a trade agreement with Europe, the issue represented more than any terms contained in the agreement—it signaled the future direction of Ukraine: to be a European democracy or to remain within the sphere of Russia.2 The brutal war to come saw its origins in this turning point. Further back in history, Japan joined the founders of the International Labour Organization in 1919 as part of its foreign policy goal to follow world trends and achieve major power status as a permanent member of the League of Nations, while the government expressed grave concerns about the possibility of foreign imposition of restrictions on labor policies. Neither the trade interests of Ukraine nor the labor policies of Japan would account for their membership choices. Through membership in international institutions, they sought to shape their association with other states. States join organizations as much for the benefit of status gained by their association with a particular group of states as for the need to cooperate on a particular set of policies. Ending membership represents the ultimate sanction because it breaks all association with the other actor. Russia’s invasion of Ukraine provoked proposals to exclude it from international organizations—in 2014, after the first invasion, accession talks for Russia to join the OECD were put on hold, and Russia was removed from the G8; the second invasion in 2022 led the OECD to terminate the Russian accession process, while the Council of Europe voted to end Russia’s membership in the organization. An emergency session of the World Tourism Body was called to hear a proposal to expel Russia from its 2. Many inside and outside of Ukraine portrayed the decision as putting its reputation at stake even when the actual levels of economic integration would not have changed in significant ways. Rejection of the association agreement in September 2013 triggered street protests and shook investors’ confidence in Ukrainian sovereign debt (Gray and Hicks, 2014, pp. 331–332).

6 chapter 1

membership, and some called for suspending its membership in the International Monetary Fund (IMF), World Bank, and World Trade Organization (WTO). These were unusual requests, as states rarely ask for and even more rarely succeed in expelling other states. In contrast, offering membership entry as a sign of solidarity happens more often. After years of deflecting Ukraine’s wish to enter the EU, in response to Russian aggression the President of the European Commission Ursula von der Leyen proclaimed in February 2022 that “Ukraine is one of us and we want them in EU.”3 Seeking to enhance its ties to Europe, Georgia filed its application to join the EU days after the Russian invasion of its neighbor. In reality Georgia and Ukraine have years before membership, but just the act of applying offers a symbolic step to deepen the association with Europe. Belonging to institutions has a significant impact on behavior. Empirical studies assess the importance of specific institutions in shaping outcomes within the issue area through comparison of policies of members and nonmembers.4 Scholars examine how the number of memberships across international organizations more generally can shape trade or conflict (Russett and Oneal, 2001; Boehmer, Gartzke, and Nordstrom, 2004; Ingram, Robinson, and Busch, 2005). Bearce and Bondanella (2007) find there is significant convergence of interests among states that share common IGO membership. Theories of international institutions explain demand for cooperation that arises from the interest in the issue area regulated. This suggests that membership rules would be clearly defined, whether by laying out eligibility for a universal organization or by establishing rigorous conditions for a restrictive organization. To maximize cooperation, accession would follow a review of whether a state meets the provisions in agreements. Members could conduct this review themselves or delegate to a committee or bureaucracy. Therefore it is puzzling that the rules of membership are discretionary and flexible. In a comprehensive review of charter provisions in international organizations, I find that a surprising number of international organizations are quite vague about membership criteria. Founding members sign up without any review process, while those who join through enlargement may negotiate terms with members but typically do so absent formal guidelines. Club IGOs 3. Euronews, 28 February 2022, available at https://www.euronews.com/2022/02/27 /ukraine-is-one-of-us-and-we-want-them-in-eu-ursula-von-der-leyen-tells-euronews, accessed 17 March 2022. 4. For example, see literature on WTO (Rose, 2004; Gowa and Kim, 2005; Goldstein, Rivers, and Tomz, 2007) or environmental policies (Young, 1999; Breitmeier, Underdal, and Young, 2011).

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 7

steer a middle course between universality and precise entry conditions. Choosing flexible membership rules supports discriminatory membership. Such institutional design provides maximal discretion to change the terms of membership as part of strategic bargains. Governments have opted to allow themselves discretion over membership in a wide range of institutions from NATO to the WTO. This discretion can be applied to lower the entry bar for some states and raise it for others. Within these organizations, geopolitical alignment strongly predicts membership patterns. As a club IGO, the OECD illustrates wide discretion over membership. While the organization is often viewed as an exclusive grouping of rich industrial democracies that advocate liberal economic policies and business regulatory standards, the reality is more complex. The OECD counts Turkey as a founding member, welcomed Mexico in 1994 before it had either completed its democratization process or achieved high income status, and most recently admitted Colombia to become its thirty-seventh member. Close affinity with the United States and Europe matter as much as economic policies to explain such decisions. In contrast, since expressing interest in the organization in the mid-1990s, Russia’s effort to join was first slowed by reviews of problems in its banking sector and corruption, and then halted in response to the invasion of Ukraine. Charter provisions are even more sparse when it comes to expelling a member from the organization. One might expect that states would enforce compliance through provisions for suspending or even terminating membership of repeat offenders. Designing terms to end a contract in case of violation would strengthen the commitment device of joining the institution. But states do not adopt this strategy of conditional membership. Few international organizations have terms for expelling a member based on noncompliance. As EU negotiators confronted the Greek debt crisis in 2010, the concern about Greek exit missed the point that while Greece could vote itself out of the union, the other EU members could not vote to throw it out! States can act outside of treaty rules, of course, but they do so in order to ostracize a pariah state rather than to enforce rules. For example, the majority of members in the Universal Postal Union (UPU) voted to expel South Africa as an expression of opposition to Apartheid policies, and they did so despite objections that there were no rules in the organization for expulsion and no evidence that South Africa had acted contrary to IGO principles for postal service cooperation. The area most likely to witness suspension is democratic backsliding, not regulatory violations. And even then, democratic backsliding by strategic partners is usually not targeted (Borzyskowski and Vabulas, 2019a).

8

chapter 1

Individual country departures correspond to a geopolitical rift. The World Health Organization (WHO) has been the center of controversy in superpower rivalries on numerous occasions—the USSR left the World Health Assembly (WHA) in 1949 to protest what they said were unfair actions by the WHO to withhold medical supplies from Eastern Europe (Fee and Brown, 2016). President Trump dramatically announced he would exit the organization amidst a pandemic because he saw the organization as favoring China.5 Cuba departed the IMF within five years of its revolution over a disagreement about repayment of loans by the Batista government that had just been overthrown. Others have had disagreements with the IMF over loans, but without exiting, though geopolitical tensions may impede the willingness of governments to work through a deal.6 In their analysis of 200 exit cases, Borzyskowski and Vabulas (2019b) show that foreign policy affinity among members reduces the likelihood of exit. Viewing international institutions as clubs embedded within broader political relations among states explains the rarity of expulsion. Just as governments do not revoke citizenship as punishment for criminal actions, most IGOs do not include expulsion as a tool for enforcing compliance. Political selection to get into international organizations magnifies the significance of expulsion threats. Not only would it deny the benefits of cooperation, but expulsion would also constitute a rejection of continued association and removal of status within international society. States would not be willing to confront a threat of expulsion for failing to meet criteria that were not necessary for entry. Such provisions would be incompatible with the entire notion of joining a community. This book will develop a theory to explain why states use club membership design for accession and how they take advantage of the discretion over conditionality and participation mandate to inject geopolitics into their membership decisions. The vague terms and room for exclusion at the time of entry represent a flexibility mechanism in the institutional design. Ad hoc criteria facilitate statecraft that uses IGO membership as carrot and stick to advance purposes beyond cooperation on the issues regulated by the rules.

5. The emergence of leaders seeking to end isolation brought the governments back. Nikita Khrushchev sought “peaceful co-existence” with the United States, and as part of this strategy returned to active engagement in WHO in 1955; President Joseph Biden terminated the process of U.S. withdrawal from the WHO. 6. The Castro government went on to repay the loans without rejoining the organization (Boughton, 2016, p. 5).

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 9

1.1 Defining IGO Membership Membership politics are the process by which states set boundaries around a community of states for cooperation. At the level of self-governing villages, the first design principle of cooperation is setting boundaries for the group (Ostrom, 2000). Global governance also requires community. When going beyond the village, community cannot be taken for granted. Nationalism looms large in the community-building process to form state boundaries. Anderson (2016) established the paradigm to view imagined communities that develop when individuals become conscious of belonging with a particular political group. In Anderson’s societal argument, the development of capitalism and the printed word opened horizontal ties of shared experience that facilitated identity aggregation. At the international level, states must also set boundaries and build communities to facilitate governance. They do so through membership in international organizations. An international organization can serve as a forum to gather states together and a resource to facilitate cooperation. An IGO can become an actor that empowers transnational bureaucrats for independent action (Hurd, 2021). Through each of these processes, the IGO builds on relational ties among states and further deepens those ties. The gains of cooperation provide incentives for states to limit joint action to subgroups. Interactions among states create the coherence for shared identity that determines who is included within these boundaries. The commitment to ongoing interaction differentiates membership from simple treaty commitments. Governments bind themselves together as a group and not just as independent actors agreeing to specific terms about behavior. Hooghe, Lenz, and Marks (2019, p. 15), in their theory of international organization, describe the nature of an evolving cooperation project among members as part of the sociality of incomplete contracting that occurs in international organizations. They emphasize that “participants are not merely making a bargain. They are also consenting to an iterated process of negotiation as circumstances change.” The layering of new commitments will often occur at a lower level of approval, without a return to the same domestic ratification process. This book examines formal intergovernmental organizations (IGOs). These are organizations that are established by three or more states and have a permanent headquarters, with regular meetings among member states. Their founding documents include membership criteria and obligations. The set of formal IGOs encompasses organizations with little institutional structure,

10

chapter 1

such as the Association of Southeast Nations at its formation in 1967, and those with more of an institutional “footprint,” such as the United Nations. It also includes organizations with a narrow focus on a single issue, such as OPEC, and those with a broader scope, such as the OECD. International organizations sometimes include more complex structures to accomplish tasks, such as emanation organizations or subcommittees. These secondary levels of organizational structure are not the focus of analysis here. In most cases the membership in the primary organization corresponds directly to membership in the subordinate organization. Informal organizations also represent a critical arena for diplomacy that is subject to socializing effects (Vabulas and Snidal, 2013; Roger, 2021). But this lies outside the scope of this study. My focus lies in explaining the puzzle of informal practices that seep into formal structures. What counts as membership? There are many forms of interaction with an IGO that can fall short of membership but nonetheless represent meaningful diplomacy. For example, observer status is widely used by the United Nations and the OECD. The GATT allowed a range of ad hoc roles that have been considered de facto membership in some studies (Goldstein, Rivers, and Tomz, 2007). This book will focus on formal membership, while considering the intervening steps that may precede a country’s joining the organization. Informal participation can ebb and flow, with little observable indication to outsiders.7 More importantly, there is a substantive difference in the commitment level of a country that chooses to attend meetings and its decision to formally commit to abide by all requirements of membership, including budgetary support and rule compliance. Formal membership brings voting rights and governance authority that implicates status as an equal actor. There is a demand and a supply side to membership. First, a government must seek to join.8 What motivates states to seek entry? There have been occasions of coercive pressure, such as Russian threats and bribes to induce former Soviet Republics to join the Commonwealth of Independent States. But at some level, sovereign states choose whether to seek membership. Not all are successful. An applicant must earn approval from other members. 7. See Gray (2018) for measures of this vitality of organizations that records emergence of zombies at the IGO level where activity wanes but the organization remains. It would be more challenging to document the change of engagement at the state level. 8. Others have given attention to expansion of access for nongovernmental organizations that in some cases become members of IGOs (Tallberg et al., 2013). This book focuses on state membership in intergovernmental IGOs.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 11

This leads to the supply side of membership, whereby other states accept an applicant subject to conditions. What incentives shape entry conditions and approval? The conditions set by members can also influence outcomes. Some countries initiate membership talks but do not complete them. This book will examine the application and the approval stages for individual international organizations and country experiences. The more aggregate analysis of membership patterns, however, will examine the final membership outcome. States can become a member either as a founder of the organization or by accession. While there are important differences in the bargaining dynamic experienced by founding states and accession states, this book will treat both forms of joining an IGO as membership. At formation, the demand and supply constraints occur in a simultaneous negotiation over the IGO itself. For accession, the late joining state has less ability to renegotiate the terms of the IGO and is more likely to face one-sided conditions for entry. Nonetheless, states whose participation is critical to the IGO may be in a position to bargain for better terms. Therefore even accession states may shape the rules. There are three ways for membership to end—dissolution of the IGO, exit, and expulsion. This book focuses on the state-level decisions toward membership, which includes the latter exit and expulsion cases. Indirectly, these actions contribute to the broader trends in the evolution of the IGO population. The “death” of organizations can be seen as the cumulative exit decisions of members who have changing interests toward the issue or in their relations with each other. While geopolitical rivalry provides the core motivation for entry into the organization, rapprochement brings IGO dissolution (e.g., the Council for Mutual Economic Assistance at the end of the Cold War). For others, the emergence of geopolitical rivalry among members acts as catalyst for death, especially in the case of security IGOs (Eilstrup-Sangiovanni, 2020). Finally, it is important to acknowledge that membership can imply different levels of activity and distribution of benefits. Some members lack influence or choose not to participate (Stone, 2011; Davis and Bermeo, 2009; Libman and Obydenkova, 2013; Hooghe et al., 2017). Research on the vitality of organizations suggests wide variation in the degree to which members engage with each other and deliver expected policy coordination (Gray, 2018). Viola (2020) shows that in many cases the expansion of participation accompanies restriction of rights through procedural rules that favor incumbents. A large swathe of empirical research examines the effectiveness of individual international organizations. Debates continue on whether the trade regime increases trade flows, multilateral aid promotes development, or human rights courts

12

chapter 1

raise protection of rights. One could reasonably ask, why study membership if some states never attend meetings or change their policies? While there are certainly examples of meaningless membership, they are not the norm. At the outset at least, accepting formal rules brings an expectation of compliance (Franck, 1990; Hooghe et al., 2017). By exploring who comes to the table, this book lays an important foundation for understanding the role of international institutions.

1.2 Membership in International Relations Theory International relations theory highlights power, material benefits, and social norms as motives for states to join institutions. Examining each in turn reveals the gaps that remain for understanding membership.

Hegemony and rule creation Powerful states often take the lead role in establishment of IGOs. Writing the rules of international order promotes their interests in systemic stability and a political order that reinforces their own position (Krasner, 1976; Gilpin, 1981; Kindleberger, 1986).9 Lake (2009) argues that a hierarchy among states arises from differential power and forms the basis for social contracts that establish an authority structure between dominant and subordinate states. Institutions connect these partners more closely to work together for common interests, and membership changes such as NATO enlargement broaden the reach of U.S. hierarchy (Lake, 2009, p. 134). Lascurettes (2020) highlights how excluding rivals through membership rules is necessary to support the depth of behavior rules among the subgroup of states that support the order. What determines the threshold when the hegemon will value participation of other states more than the risk of losing control over the organization? Stone (2011) argues that informal influence allows the United States to balance competing goals for participation and control. He argues that institutions such as the IMF and the WTO induce the hegemon to follow the rules in normal times while allowing deviation over critical issues. This theory suggests IGOs will admit a broad membership on the assumption that there is differential application of the rules. Sponsoring allies and former colonies for 9. While power is the basis of hegemonic leadership over institutional creation, the specific orientation of the hegemon toward the ideas of liberal multilateralism shape their propensity to build institutions (Ikenberry, 2001).

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 13

membership represents an important side of informal politics exercised in the shadow of universalistic rules. Small states also form and join IGOs. Democratizing states fit rules to their own needs in new organizations and join at higher rates than other states as a way to provide for public demands (Poast and Urpelainen, 2013, 2018). Outside of institutions they have the least influence and their capacity constraints often prevent their achieving important governance tasks. International cooperation offers a solution to their weakness—even if power asymmetry continues within a rule framework. Gruber (2000, p. 8) contends that large states use their “going it alone” capacity to force on smaller states terms that they would not prefer over the status quo, while setting new baselines such that joining a coercive institution is still better than “being completely shut out.” Even within the constraints of power, states retain the choice to join. At the multilateral level, participation arises from a national decision and a collective decision. This requires looking more carefully at how IGOs form a club with capacity to provide benefits to members and exclude benefits from nonmembers.

Designing Rules to Overcome Barriers to Collective Action Demand for regimes arises when states would benefit from cooperation and an institution helps them overcome market failures that would prevent such cooperation. Keohane (1984) develops the core logic of functional demand for institutions based on their ability to lower transaction costs. His theory focuses on the collective action problems and asymmetric information that characterize cooperation for public goods. On the assumption that benefits are non-rival and non-excludable, institutions are necessary to provide information, monitor compliance, and link issues in ways that support cooperation among a large group. Membership itself is not a central question. For a pure public good, screening membership is ineffective because members and nonmembers alike can benefit from the cooperative output. Nevertheless, limiting cooperation to a subset of states is widespread practice by international institutions. Using a club model of cooperation builds on common interests within a smaller group of states on a subset of issues— for example, the early years of the trade regime excluded illiberal states and kept narrow focus on trade policies at the border to facilitate easier bargaining over agreements (Keohane and Nye, 2001). Trade policies allow states to discriminate in provision of market access to members and nonmembers in

14

chapter 1

the multilateral trade regime, and governments apply elaborate rules of origin to further differentiate trade among states with preferential trade agreements (e.g. Gowa and Kim, 2005; Mansfield and Milner, 2015). States rely on alliances to defend against shared threats, with careful selection of members to maximize security (Sandler, 1999). This reflects the fact that many cooperation problems constitute impure public goods where exclusion is possible. Even in the area of climate change, which represents a classic public goods issue, Keohane and Victor (2011) argue that the environmental regime complex will function better when multiple institutions such as the G8 and EU compete to find solutions to problems among smaller clubs of leading governments. In order to transform environmental cooperation into a club good, Nordhaus (2015) recommends linking climate policies to trade. He proposes using tariffs as a punitive measure against states that do not join the climate change organization. The article highlights that the key to success for international cooperation lies in finding an effective exclusion mechanism. Yet few IGOs follow this solution to expand issue scope and coerce membership through punitive sanctions. Doing so relies on positive utility from enacting sanctions for current members, and this rarely attains credibility. Even in the trade mechanism suggested by Nordhaus (2015), economic theory and empirical trends in trade agreements both contradict his assumption that all states benefit from raising tariffs. More often, such as in the IAEA or climate change protocol, issue linkage uses various forms of aid as a carrot to entice membership. In both positive or negative sanctions in a universal group and reducing membership to smaller groups, these theories recommend changing the issue scope so that cooperation can be treated as a club good for provision by a subgroup of states rather than as universal cooperation for public good provision. This leads to the question of how states design the exclusion mechanism for international institutions. To address fears of free riding and cheating, membership conditions should represent a significant hurdle such that those unwilling to comply with the rules will not become members (Koremenos, Lipson, and Snidal, 2001). Lower barriers to entry would be expected for coordination games like standard-setting, where wider participation generates more benefits. This leads to contradictory predictions for membership provisions when cooperation involves both distributional and enforcement challenges (Koremenos, Lipson, and Snidal, 2001, p. 796). The trade-off between depth of rules and breadth of participation can be considerable in the face of diverse state interests. The optimal size of an IGO depends on the enforcement concerns and the distribution of gains

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 15

from cooperation in an issue area as well as on the preferences of states (e.g. Martin, 1992; Drezner, 2007; Koremenos, 2016). A small group with similar preferences can more readily reach agreement for cooperation and faces fewer monitoring problems (Kahler, 1992; Downs and Rocke, 1995; Thompson and Verdier, 2014). This justifies rigorous screening based on performance capacity. In the context of public goods provision, however, a smaller group also means that other states can free ride as they choose not to contribute to cooperation at the high level demanded (Stone, Slantchev, and London, 2008). The substantive significance of “deep” agreements diminishes due to the limited number of participants. A larger membership gains from pooling resources and taking advantage of the economies of scale, which is the rationale for cooperation through formal organizations in the first place (Abbott and Snidal, 1998). Where international relations theories highlight distributional conflict over accession, they build the expectation for international organizations to sort states into subgroups with similar preferences on the regulated issue. Downs, Rocke, and Barsoom (1998) suggest the optimal pathway for cooperation outcomes lies in sequential liberalization, whereby small groups set the rules and gradually expand to admit new members after their preferences have converged. Gray, Lindstädt, and Slapin (2017) model enlargement scenarios in which the location of the original group and applicants on a unidimensional space determine the probability for enlargement. Small and homogenous founding groups can achieve stable enlargement without changing the organization, whereas a more diverse set of founding states may find that misperceptions about applicants lead to enlargement that changes the level of ambition in agreements. Both of these theories focus on the unidimensional preferences of members in the issue regulated by the regime. Voeten (2021) contends that the differences among states lie within a low-dimensional space that can be defined in terms of support for the Western liberal order. Viola (2020) develops a theory of how states manage diversity through a strategy of assimilative multilateralism, with entry conditional on conformity within a specific range of issues. These theories follow the logic that similarity supports cooperation, but leave open the question of how states coalesce around similar interests within the regime issue area or broader world order. The institutional design theories reviewed here highlight the importance of studying conditions for entry. Yet they cannot answer why so few IGOs apply rigorous screening. Letting in noncompliers lowers cooperation while adding to the burden of high cooperation states.

16

chapter 1

The Social Role of Organizations States may look outside of the policies regulated by the regime when considering the benefits of collaboration with particular states. Association with other actors through organizational membership carries spillover effects. Joint membership forms an association that shapes investor perceptions and security coalitions (Gray, 2009; Brooks, Cunha, and Mosley, 2015; Henke, 2019). Following the foundational work of Bull (1977), I examine international society from a perspective that heeds both power and the dimensions that are based on a social process of interaction and shared community norms. Hooghe, Lenz, and Marks (2019, p. 2) make the critical point that “we need to consider how participants feel about being bound together in collective rule” through studying international governance as it serves functional and social purposes. They compare market traders to those bound in marriage to illustrate how the different perceptions of community could influence cooperation. This turns us to the question of group formation. Absent dating services, how do states form the right group with a sense of community? Membership decisions account for the probability of cooperation by the new entrant and the expected gains from establishing a closer association. When members care about both the shared good and the attributes of other members, they may devise “discriminatory clubs” that select according to the desirability of the applicant and not just their expected contribution to cooperation outcomes (Cornes and Sandler, 1996, p. 385). As an example, members in a social club care not only about the entertainment activity itself but also about member composition. This changes the logic of cooperation. In most models of membership in institutions, the bargaining problem models anonymous states holding proportional contribution to cooperation based on their size and interests (e.g. Stone, Slantchev, and London, 2008). In discriminatory clubs, where the attributes of other actors are important to members, the utility of joining a club consists of two components—provision of goods by the club, and consumption of characteristics of other members (Cornes and Sandler, 1996, p. 385). When members care about who joins, there is “nonanonymous crowding” in the provision of the club good. The common example is a golf club, where members value association with those of high socioeconomic status as opposed to simply caring about the number of people using the course or their golf abilities. There are several reasons why governments should care about which states they form ties with through IGO membership.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 17

First, membership defines legal status. Belonging to organizations forms the basis of a state’s position in international society. Recognition by other states is necessary to establish international legal sovereignty, but there are no consistent rules guiding such recognition (Krasner, 1999, p. 15). Indeed, research on diplomatic recognition, which is the most basic status requirement, indicates that relational variables matter more than country attributes (Kinne, 2014; Duque, 2018).10 In addition to diplomatic recognition by other states, membership in organizations such as the United Nations is one of the most visible forms of recognition. States that may otherwise not benefit from full rights of sovereignty in the sense of independent control of their territory may nonetheless gain equality through membership, such as the case of India being a member of the League of Nations and founding member of the United Nations while it remained a British colony. Diplomatic recognition reveals network effects in which states pattern their own decisions on those of other states that are seen as friends (Kinne, 2014, p. 248). Similar kinds of dependent decisions across IGOs occur, such as when membership in the United Nations forms a precondition for membership in many other organizations. Each additional membership can reinforce the extent to which a state enjoys recognition. As distinct from the specific benefits of the IGO, through supporting international legal sovereignty, IGO membership brings diffuse benefits that include both material and normative resources, such as prestige before domestic audiences from the appearance at international venues and reassurance to investors about certainty of contracts when dealing with a recognized state (Krasner, 1999, p. 16). Second, membership creates peer groups. Through the activities of an organization, opportunities for interaction with other members increase and the norms of the organization shape behavior of members. States become identified with the organization such that the members’ own reputation can be influenced by the reputation of fellow members of the organization. Both economic and security actors may respond to this reputation in ways that generate material consequences. Organizational membership can impact country risk ratings and bond yields. Dreher and Voigt (2011) find that independently of the quality of domestic institutions, membership in international organizations improves the country risk rating. Gray (2013) identifies peer effects in 10. Social identity theories may not readily apply to the formation of collective identity among states, but state recognition provides a first step in such a process. See the discussion by Greenhill (2008).

18 c h a p t e r 1

which a state joining an organization can be seen as a better or worse investment environment simply because of its association with a particular group of states through joint membership in a regional economic organization. She theorizes that investors use the IGO membership information to inform their risk assessment, and finds evidence that bond yields change in response to accession negotiations. Third, membership establishes a forum for interaction. States build trust through social ties, which can occur irrespective of the issue area of cooperation and hold positive externalities for trade and security. The step to establishing closer relations through participation in a formal organization with joint governance creates opportunities for members both to learn more about each other and to form in-group identity. Indeed, Ingram, Robinson, and Busch (2005) find sociocultural IGOs promote trade as much as economic IGOs. Yet the quality and degree of interaction may vary depending on the institutional context and relative status of each additional tie for groups of states. Boehmer, Gartzke, and Nordstrom (2004) find that membership in structured IGOs with formal rules and procedures reduces the probability of conflict onset, while Hafner-Burton and Montgomery (2006) emphasize IGO membership holds pacifying effect conditional on how the IGO determines the social network position of a state. Taking a network approach to examine convergence of IGO membership, Kinne (2013a) shows that deepening ties across IGOs reduces the probability of militarized disputes. This could arise through states screening out conflict-prone states when making membership decisions (Donno, Metzger, and Russett, 2015). States that associate together in IGOs send informative signals about their type to both potential investors and disputants. Across each of these dimensions, the attributes of other members shape the benefits of shared membership. States will want to be selective about whom they recognize as equal and associate with as a peer for greater interactions. They hold incentives to favor entry for states with desirable characteristics. Criteria for discrimination in social clubs range from income to race, but it is unclear what would be the equivalent criteria for desirable attributes of states within international society. Region, economic development, prestige, and security alliances all form potential bases for discrimination. Internal politics may elevate one dimension over another. The promise of gains from status by association makes membership more attractive. A government could even decide to join an organization that offers few benefits from the provision of goods but yields large status gains from

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 19

association with a particular group of states. Classifications within organizations take on larger significance because of their role of sorting states into groups.11 At the same time, members may screen out those who offer less benefits from association even when they could otherwise meet performance conditions for compliance with rules. This can be seen from data on networks of diplomatic recognition, whereby states cue off of central states’ decisions more than of those on the periphery (Kinne, 2014). States have reason to fear that they will be branded according to joint association with other states. Johnson (2011) argues that there is a process of guilt by association, whereby unfavorable views toward one member can lead to overall skepticism toward the organization as a whole. The perception that the hostile state holds institutional or ideational influence gives rise to such negative feelings, which is supported by evidence from public opinion polls. Gray (2013) further demonstrates that the strength of association with other states can have unexpected spillover; membership in a regional organization will impact bond yields, as members gain better reputation from joining an organization with low risk countries and suffer when joining an organization with high risk countries. Group membership is how international society allocates status across states. But rather than being a uniform and objective metric, status by association varies according to the criteria used as the exclusion mechanism in group formation. Dafoe, Renshon, and Huth (2014, p. 375) define the concept: ‘Status’ is an attribute of an individual or social role that refers to position vis-a-vis a comparison group; status informs patterns of deference and expectations of behavior, rights, and responsibilities. Status categories may be dichotomous (e.g. membership in a group) or rank based (e.g. position in a hierarchy). A change in an actor’s status implies a change in at least one other actor’s status, either because of a change in rank or because of a (perhaps slight) change in the meaning of membership in a group. The connection to membership is clear—status is a function of community recognition, and joint membership in organizations is one visible signal of such recognition. While diplomatic recognition is the baseline of status, membership in organizations offers a more differentiated perspective. Paul, Larson, 11. In one such example, states seek to “graduate” from World Bank lending programs despite the prospect of less foreign aid, and undertake more political reforms for this goal (Carnegie and Samii, 2017).

20

chapter 1

and Wohlforth (2014, p. 7) define status as “beliefs about a state’s ranking on valued attributes,” such as wealth or diplomatic clout, and note that status manifests itself in international politics through “membership in a defined club of actors.” Elite clubs like the G7 or the permanent five members of the UN Security Council are considered status markers. The pursuit of status has long served as motivation of states. From Thucydides to contemporary scholarship, theories account for wars fought in the name of honor and prestige. Keohane (2010) advocates “esteem” as a source of incentives for governments to support climate change policies. Although there are different nuances across terms, one can reasonably aggregate honor, prestige, and esteem as forms of status. Furthermore, the rival nature of status can also be understood through its relative position. If all states are members of an organization, membership no longer confers status to one state relative to another. Finally, the status by association in IGO membership is differentiated by the quality of other members rather than simply by an attribute of one state. Notably, status as a motivation for entry into organizations differs from both the socialization that occurs after joining and scripted behavior. Johnston (2001) argues that institutions represent a social context in which shaming and backpatting influence states toward compliance. This socialization process, however, depends on the state having first joined the organization. To the extent only pro-social states choose to join the organization and are allowed to accede by members, constructivist theory about socialization in IGOs is equally subject to the selection bias concerns that face functional theories of institutions. In his study of China’s decision to engage with international institutions for security cooperation, Johnston (2007) argues that concerns about avoiding isolation and taking conformist positions were more important than the impact of proposed commitments on relative power. In his theory, the first step occurs as mimicking, when states follow the behavior of others, based on the assumption that if others are joining the institution it must present benefits. After entry, states are then socialized through interaction as they respond to social rewards and punishments and build their own internal organizations to support their work within the institution. A government that once was content to be isolated, through membership becomes attuned to its position within the group in ways that make it susceptible to social pressures and lock in a new direction of cooperative policies. But the theory does not answer the question about why states choose to mimic one group of states and not another.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 21

Placing social interaction at the center of theories, constructivist international relations scholarship offers several insights related to membership in organizations. In constructivist theories, the historical context of relations among states and the decision processes within states exhibit path dependence and cognitive biases that differ from simple calculation of interests. Organizations may serve as an arena for rule-bound behavior that follows a logic of appropriateness (March and Olsen, 1998). From this perspective, states join organizations because that is what states do, irrespective of the costs and benefits of specific organizational membership decisions. Current members find themselves caught in their own rhetoric of inclusion and cooperation, and are unable to turn away applicants (Schimmelfennig, 2001). Variation in membership patterns could arise through channels of emulation, whereby states follow the IGO membership decisions of influential regional leaders. When states join organizations as part of a script for modernity to follow the behavior of other states, membership in organizations takes on larger purpose than the simple provision of benefits. As described by March and Olsen (1998, p. 964), organizations like the OECD or European Union are “creators of meaning in general and more specifically of identities.” Through joint membership, deep ties emerge that could influence state preferences. Constructivist theory emphasizes the possibility for states to set aside self-interest within a larger collective identity (Wendt, 1994). This collective identity emerges from repeated interaction. These theories of socialization acknowledge the importance of forming ties in an IGO. Next we consider how security underlies the choices about which states seek closer association—the choice of whom to mimic and what attributes to value in cooperation partners arises from geopolitical alignment.

1.3 Geopolitical Alignment as Basis for IGO Cooperation States take sides within international politics. The shorthand term for this is geopolitical alignment, which refers to a like-minded approach to world affairs and especially to international security problems. The concept overlaps with alliance structures but can differ in important ways. States ranging from Switzerland to Israel fall within a sphere of common security interests with the United States while never having established alliance ties. The states of Southeast Asia have formed a like-minded approach to security that emphasizes non-intervention in domestic affairs even while their alliance

22

chapter 1

affiliations differ. During the Cold War, the nonaligned movement countries engaged in security cooperation by the joint decision not to become allies with either the United States or USSR. The concept of geopolitical alignment differs from that of ideology because of the defining role of security. Whereas liberal ideology differentiates between democratic and authoritarian regimes or market and non-market economies, geopolitical alignment supports cooperation with any political regime type. Likewise, ideology prioritizes differences in economic policy orientation that are not significant from a security standpoint—partnering with states that uphold liberal markets or with those with more intervention for developmental or socialist policies may serve security interests despite vast differences in ideology. Two features of geopolitical alignment make it the favored selection criterion in the politics of joining international organizations. First, alignment offers information about reliability. Second, it aggregates interests to support bargaining. The former reduces fears that a prospective cooperation partner will cheat while the latter encourages a broader view of the distributional gains from cooperation. Here I briefly take up each in turn. As an information tool, geopolitical alignment provides a valuable cue about the quality of cooperation expected from another country. Selecting partners for cooperation requires an assurance about future behavior. States that change regulatory policies or pool resources expose themselves to risk if the others fail to comply. But since compliance types are difficult to judge ex ante, states must seek information from other sources. The accumulation of security cooperation reflected in activities ranging from alliances to military training exercises and joint foreign policy statements or voting in the United Nations provides many opportunities for states to learn about the geopolitical alignment of prospective partners. Their success coordinating on issues related to security builds trust to support subsequent cooperation on new challenges. Equally importantly, geopolitical alignment expands the bargaining range to include more issues. States that share interests for foreign policy can more readily generate mutual gains from trade-offs between economic and security policies. More expansive cooperation is possible when linkages support sharing economic gains. This can facilitate bargaining even beyond the narrow sphere of allies and for cases where the economic exchange is asymmetric. States may tolerate cheating or unequal distributional outcomes within the confines of the regime in exchange for wider benefits across the relationship.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 23

The role of shared security interests, to provide information and linkage channels, differs from the conventional view in the literature on cooperation. In his foundational theory of international regimes, Keohane (1984) focuses on the role of enforcement and issue linkages carried out within the regime jurisdiction as ways to overcome the information asymmetry that hinders cooperation. Instead, my argument highlights how political relations that exist prior to and outside of the regime provide information about expected compliance. When states let non-regime issues such as security determine membership choices, they open themselves to less effective regimes through overexpansion, as they let in unqualified applicants, or underprovision of cooperation, when they leave out otherwise qualified entrants. In the leading realist perspective on overlapping security and economic interests, Gowa (1994) theorizes that the security externality of trade motivates allies to trade more with each other in order to share the income gains from economic exchange.12 But this logic applies to the relationship between pairs of states where common security interests are certain, such as bilateral trade between allies during the Cold War. It cannot explain the surge of institutionalized cooperation after the end of the Cold War, when there is less certainty about which states will be allies or adversaries. The puzzle remains of why states would commit to multilateral cooperation when their security relationship could change. Indeed, within the context of the long-term commitment to repeated action in an IGO, a security externality could worsen the bargaining problem by increasing distributional stakes (Fearon, 1998). The security linkage that underlies multilateral institutions offers an alternative logic to one based on principled beliefs. In his theory about ideology and multilateralism, Voeten (2021) argues that the United States uses multilateralism to advance its ideological principles. In order to move the status quo in the preferred policy direction, the United States coerces those joining institutions to follow regime rules as part of a strategy to diffuse liberal principles of free tade and democracy. Screening for shared beliefs at entry and upholding high compliance with policies would advance those goals. But security linkage has no such restrictions—patronage politics to favor allies could even motivate states to lower standards for entry and compliance. Indeed, U.S. 12. Gowa (1994) shows that trade gains form a positive security externality when allies trade, in contrast to the negative externality arising from trade between adversaries. The bipolar structure of an international system supports open markets more than a multipolar system because the security externality motivates altruism between allies in their trade relations (Gowa, 1989).

24 c h a p t e r 1

allies have been shown to undertake less economic reform when entering the GATT or receiving IMF loans than their counterparts (Stone, 2008; Davis and Wilf, 2017). It is difficult to differentiate between the ideological and security logic. Shared values and beliefs about the organization of society form a foundation for cooperation on security, which contributes to the overlap between geopolitical alignment and ideology. During the Cold War this overlap occurred in the bipolar division of competing alliances between communist and capitalist sides. One might expect that the role of alliances in shaping entry into multilateralism would end after the Cold War, along with the decline of the ideological basis for alliances. But it has not. States are still more likely to enter IGOs with their allies. The security logic of geopolitical alignment to build a coalition through multilateralism remains amidst the uncertainty of a changing order. Joining international organizations together strengthens security coalitions. In hierarchical relations of exchange, states reinforce their ties through offering side payments to support security cooperation (Lake, 2009). This process is easier when states share membership in organizations so they can exchange favors on priority issues (Henke, 2019). As fellow members in an IGO, states can use patronage or bribery to gain leverage over critical swing states in a broad security coalition. Research confirms that within multilateral economic institutions, allied states lend more and trade more with each other than with other members (Thacker, 1999; Gowa and Kim, 2005; Stone, 2008; Dreher et al., 2013). Policy coordination outside of defense policies also helps states to signal intentions of goodwill and commitment to security partners (Morrow, 2000). Henke (2019) shows that diplomatic embeddedness through the exchanges that take place in multilateral fora support the formation and maintenance of multilateral military coalitions. Linking economic and security cooperation has also been shown to reduce conflict within alliances and increase alliance performance (Powers, 2004; Poast, 2013). Joint association in an international organization sends a message of solidarity. At the same time, excluding rivals denies them the benefits of the organization. States have less leverage to punish a rival after it has joined a multilateral organization. Indeed, according to Carnegie (2014), rivals receive a large boost to their trade after entry into the trade regime. Multilateral norms do not prevent biased allocation within organizations, but they make it more difficult to restrict access to a state than if it were a nonmember. Exclusion of rivals also signals outsider status by isolating them from routine diplomatic exchanges.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 25

The status benefits of IGO membership also engender cooperation among allies. From basic recognition of sovereignty to major power status, IGO membership converts relationships into a broader standing within international society. Reputation may generalize across members in the institution. This can deliver additional benefits as states improve their standing in the eyes of investors or gain credibility vis a vis hostile states. Keeping these benefits within a security community is optimal. The social interaction amplifies the preference for association with security partners. Choosing to form closer relations by joint membership arises from willingness to engage in close interaction and share cooperation benefits. Joint membership also informs all states about the social categories of which states work together. States will value status by association with security partners more than status by association with other states. In the domestic politics of international cooperation, accession offers a window for security to take priority. Diplomats and top leaders take charge of treaty negotiations. Given their mandate over foreign policy, these actors have incentives to allocate significant attention to security stakes and diplomacy (Bueno de Mesquita and Smith, 2012; Milner and Tingley, 2015). As these negotiators coordinate with domestic actors over treaty ratification, they can emphasize geopolitical alignment with other states to build support outside of the direct constituencies for the agreement. Policies within the regulated issue areas confront stakeholders lobbying for gains and resisting costly adjustment. Issue linkage offers an effective tactic for breaking through divisions. In this case, joining organizations helps diplomats and leaders with strategic goals frame the broader stakes in cooperation and avoid zero-sum single-issue politics. The public may favor cooperation with allies independently of the specific issues in the treaty (Carnegie and Gaikwad, 2022). Adding potential gains from the security dimension makes entry more likely than if partners are seen as rivals or the IGO is evaluated only in terms of the issue area. From a coalition-building view of IGO membership, states seek additional leverage in their relations with other states by broadening and deepening their sphere of connections through IGO membership ties. As a form of balancing, strategic use of IGO membership represents a useful tactic for both strong and weak states. It also remains relevant during periods of low certainty over alliance relationships, which calls for keeping options open on whom to be able to influence. This supports an expectation that many states will design IGOs as discriminatory clubs and use that flexibility to favor their security partners. There will be less variation by issue area or distribution of power

26

chapter 1

than implied by competing explanations focused on cooperation problems and distribution of power. To evaluate the argument, the book will test three core claims about how states discriminate over membership. The first hypothesis addresses the rules and decision-making procedures that form a background condition for discrimination. The second hypothesis looks at the nature of bias that shapes membership patterns. The third considers how the entry process allows for differential costs of entry. Hypothesis 1, Discretionary Rules: States will design IGO accession rules to provide discretion over selection. More international organizations will follow the design of club models, with an exclusion mechanism based on voting rather than policy evaluation of rule compliance. At the design stage, it is difficult to define ex ante the in-group, and states want the flexibility to engage in discriminatory practices. Therefore, they leave the entry qualifications vague while assuring control over selecting who gets in. The selection could favor anything—security, culture, economic interests, or compliance—because the rules do not specify. The second hypothesis explains the geopolitical logic that drives the pattern of membership in IGOs and is the primary focus of the book. Hypothesis 2, Geopolitical Discrimination: States with shared geopolitical alignment form organizations together and are more likely to join the same organizations. In this relational theory, states discriminate to favor others based on their preexisting security ties. Shared alliances and similar voting in the United Nations serve as proxies for measuring like-minded orientation to security issues. Alignment with other members and not just the largest power can support entry and continued membership. Yet non-allies and even rivals can and do join IGOs—the hypotheses are probabilistic and not deterministic. As states consider cooperation partners, non-security gains may outweigh the security factors. In the sequencing of member expansion, early entry by allies will consolidate the voting core and rules, while non-allies must wait longer and do more to win approval. The third hypothesis explains the conditions for entry. Hypothesis 3, Favoring Friends: States with shared geopolitical alignment with other members will make fewer reforms as a condition of entry. This final hypothesis is nested within the first two because discretionary terms of membership allow variable conditions across applicants, and lowering the bar for friends

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 27

facilitates early entry. Holding up non-allied states until they meet a higher threshold accommodates mixed incentives about who can join—as in societal discrimination where the out-group applicant must be twice as qualified on performance basis to overcome the bias against them. The argument holds implications over the politics for ending membership as well. Geopolitical alignment supports ongoing membership after having joined an organization. The discretionary approach to accession extends to rules for exit and expulsion. States are not required to prove compliance in order to remain in the club. Empirical analysis will compare the role of geopolitical alignment with the demand for membership based on interests within the issue area regulated by the IGO. The large body of literature on functional theories of institutions posits that the desire to achieve mutual gains in the face of market failure motivates cooperation in international institutions; this leads to the baseline expectation that interests, information, and policies within the issue area should explain who joins. Since geopolitical alignment and interests within the issue area overlap entirely in the area of security organizations, looking outside security organizations is necessary to test the hypotheses. In the area of economic organizations one can compare how economic interests contribute to expected benefits from membership relative to the impact of geopolitical alignment. Supporting evidence would include examples where states that have little engagement in international trade join the trade regime at the encouragement of an ally. Further evidence would include internal statements from diplomats and security hawks in support of joining the IGO, and lenient provisions to allow entry without requiring substantial economic policy reforms. Raising non-trade issues and extra concessions as a condition for membership to block entry by a rival state represents the exclusionary side of discrimination. In contrast, support for functional theories would include evidence that most of the variation in membership arises from differences in the trade interests of states. The domestic political interests would largely revolve around debate among different economic ministries and business groups, and accession negotiations would extract substantial concessions to ensure conformity with rules. There would also be divergent expectations about the conditions that would lead states to exit or be expelled from an organization. Whereas the discriminatory club model of IGO membership suggests exit would be largely independent of compliance with IGO rules, the functional model of institutions implies that exit and expulsion would occur after a period of noncompliance.

28

chapter 1

Although the argument implies that states expect to deepen cooperation with other states through shared membership, this book will not evaluate the effects of entry. Nevertheless, there is a dynamic where states are selecting based not simply on existing security but also on the security relationships they want to develop by means of shared membership. The politics of engagement with Eastern Europe or newly independent states during the Cold War stand out as such examples. At the two ends of the spectrum of geopolitical alignment, there are clear incentives to favor allies and exclude rivals. But for swing states there may be differences in approach, as diplomats are making bets on the future trajectory of a state and trying to shift that trajectory. The hypotheses above can only partially capture this strategic dynamic: flexible rules maximize discretion for borderline cases and on average those that are closer in geopolitical alignment will be more likely to join, with lower conditions. Many of the hardest cases to predict will be those where contextual factors will shape perceptions of whether geopolitical alignment of the state is in transition.

1.4 Chapter Overview The book proceeds to further develop the logic of the argument and explore the patterns of membership politics. Moving from the general theory of international organizations as discriminatory clubs to applications in the context of specific international organizations and country experiences, I will balance conceptual approaches with the nuance of history and mixed motives. Where the aggregate analysis of international organizations uses proxies to measure geopolitical alignment, case studies will probe more deeply the various forms of in-group identity on security issues and the interaction between domestic politics and international cooperation. Although unable to leverage exogenous shocks or randomized experiments for rigorous identification of causal effects, the sum of descriptive inferences across mixed methods analysis supports the hypotheses and builds an agenda for future research. The central research question of the book asks how states choose their cooperation partners. This introductory chapter has laid out the core claim that geopolitical alignment shapes multilateral cooperation on non-security issues through membership politics. When international organizations become discriminatory clubs, they set the boundaries for cooperation among a subset of states chosen for their affinity. Security forms the basis for affinity in this study, but the larger claim contends that like-minded states cooperate

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 29

by means of biased membership decisions. Viewing membership as a choice of association within international society differs from the conventional expectation that international organizations represent contracts to uphold the interests of powerful states and coordinate common interests within an issue area. International organizations are a heterogenous set of institutions that vary in size, rules, and issue mandate. The research design of the book takes different approaches to assess geopolitical alignment relative to other conditions when controlling for some of the features that differ among organizations. Chapter 2 compares the design of all IGO charters, looking at how discretion appears across IGOs in different issue areas and with different membership size. The empirical analysis of membership decisions in chapters 3–5 focuses on economic IGOs to evaluate whether security interests unrelated to the mandate of the organization emerge as a factor over membership alongside economic interests. Chapter 6 looks from a country perspective at membership decisions across all organizations over time. Then the book turns to IGOs with two different types of entry rules: chapter 7 focuses on organizations with a regional focus, while chapter 8 looks at those that explicitly embrace the principle of open eligibility for all states. Across this range of institutional settings, geopolitical alignment emerges as a consistent factor in membership. This is not limited to the domain of global institutions or those within Europe; organizations large and small and those across different regions heed the pull of geopolitical alignment when considering members. Yet other factors also matter, and statistical analyses with control variables and case studies with attention to different narratives reflect on these mixed motives. Chapter 2 develops the theory of membership that makes IGOs form discriminatory clubs. States seek both gains from cooperation and status from association with other states. This dual purpose explains the form of membership provisions that are designed to promote cooperation among a community of states. The central importance of geopolitics motivates the prevalence of discretion over member approval. States retain control to choose with whom they cooperate. The chapter goes on to test the hypothesis about discretionary design of membership provisions with a comprehensive analysis of IGO charters and their terms for membership selection. A typology of IGO membership provisions illustrates variation across the dimensions of participation mandate and conditionality terms. These concepts are mapped onto accession rules for 322 international organizations using a new dataset. For issues that represent

30 c h a p t e r 1

global public goods, universalistic principles underlie membership with open eligibility for all states. Most join easily. Nevertheless, screening for entry into universal organizations determines who counts as a state within international society, and so even universal organizations include some selection. Attention is given to how design varies by issue area, but there are surprisingly consistent features across issues. For the majority of issues that constitute club goods where exclusion is possible, states design IGO membership provisions to select a smaller group of states. Surprisingly few organizations screen for compliance based on objective performance standards or policy review. Instead, vague eligibility terms, negotiable terms, and the requirement of member approval characterize club standards for membership. I argue that states choose this design structure in order to maximize their flexibility to have informal norms and geopolitical interests operate as de facto criteria for who joins. Alongside the discretionary approach to accepting new members, IGOs rarely terminate membership over noncompliance. This chapter explores why states refrain from following reciprocity strategies that would call for threatening to expel states that repeatedly reject IGO rules. Analysis of IGO charters examines which types of IGOs include provisions for expulsion and exit. The infrequency of member suspension for noncompliance upholds the logic that membership in an IGO confers a form of citizenship within society rather than a simple contract. Chapter 3 evaluates the second hypothesis that geopolitics correlates with membership.13 Looking at multilateral economic IGO membership offers a sharper test, since the substantive focus of the organization itself does not require coordination of security policies. We use data on alliances and UN voting similarity to measure geopolitical alignment, and compare the geopolitics hypothesis with the benchmark model that organization membership reflects economic interests measured by the trade ties between countries. Analyzing membership patterns for 231 multilateral economic organizations from 1949 to 2014, we use a finite mixture model to examine the relative importance of economic and security considerations, finding that geopolitical alignment accounts for nearly half of the membership decisions in economic institutions. The geopolitical origins of IGO membership represent an important mechanism connecting the security and economic behaviors of states. 13. This chapter is based on an article co-authored with Tyler Pratt, “The Forces of Attraction: How Security Interests Shape Membership in Economic Institutions,” Review of International Organizations 2021.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 31

Turning to focus on one international organization, chapter 4 looks at how the trade regime expanded from a small club to a nearly universal organization.14 The trade regime is important for both its substantive significance and because archival records include information on all countries that have applied for membership, which is often not available for other organizations. This allows comprehensive research into the full accession process for all eligible countries over the period 1948 through 2014. Using a duration model to analyze when states apply and the negotiations for accession offers insight into the demand and supply sides of membership. Consistent with hypothesis 2, geopolitical alignment predicts who applies to join, and consistent with hypothesis 3, corresponds with lower cost of entry, as shown by shorter accession negotiations with fewer concessions. Our findings challenge the view that states liberalize first in order to join the regime. Instead, democracy and foreign policy similarity with members encourage states to join. Moreover, the level of tariff cuts and conditions for accession are higher for non-allied states. While now there is much attention to China’s role as a challenger within the regime, from a historic and comparative perspective, China was late to join and made substantial concessions as a condition for entry. The OECD exemplifies the discriminatory role of IGOs. Yet its exclusive nature is not as its nickname “rich country club” would suggest. Chapter 5 documents how this organization plays a key role in shaping international coordination of regulations on policies that range from taxation and investment to education and corruption, but also constitutes a quasi-alliance bringing together “the West.” Scholars and commentators alike refer to OECD states without sufficient attention to the unifying principles that shape who can join this exclusive international club. The organization provides both public goods in the form of policy information and club goods in the form of status. Through a process of self-selection by applicants and screening by members, the organization has managed gradual expansion while preserving its value as an elite club of like-minded states. Informality of accession criteria has allowed flexibility to raise and lower the bar for entry. Statistical analysis highlights that democracy and geopolitics correlate with earlier entry into the OECD relative to other countries, while there are less clear patterns for the role of trade and financial openness. Case studies of Mexico, Korea, and the Czech Republic are used to examine how prospective OECD membership motivated reforms in 14. This chapter is based on an article co-authored with Meredith Wilf, “Joining the Club? Accession to the GATT/WTO,” Journal of Politics 2017.

32

chapter 1

regulatory policies and trade. These countries sought the status of association with the advanced industrial democracies. On the basis of shared liberal orientation and geopolitical alignment, they were accepted into the club. At the same time, a case study of Brazil highlights how its initial reluctance to seek OECD membership reflected a political preference to remain distant from the advanced industrial nations even as its economy and policies had become highly integrated with these states. The analysis also examines how Brazil’s effort to establish closer ties with the United States led it to recently apply to join the OECD. In chapter 6, the book shifts from the IGO level to a country-level perspective. This chapter explores the experience of Japan from 1853, when it first opened to the West, through today, when it is member of more than 80 IGOs. With attention to domestic politics and the perception of allies and rivals, this chapter closely examines why international organizations are viewed as a way to shape the national position in international society. Tracing Japanese foreign policy through its approach to IGO membership highlights the role of status-seeking through association with allied countries. Whether it be Meiji Japan’s effort to prove itself a sovereign state that could win itself free of the unequal treaties, or post-WWII Japan’s effort to reenter international society after defeat in war, IGOs have long been about more than functional tools to provide specific goods. While foreign policy goals lay behind Japan’s membership decisions, entry into organizations then formed key leverage points to bring about domestic reforms. The chapter documents how Japan’s relations with the UK during the Anglo-Japanese alliance and with the United States during the post-war alliance shaped its entry and exit from organizations, and examines how changing diplomatic relations with China have influenced the Japanese position toward China’s role in IGOs. The theory helps account for several surprising membership outcomes—such as the improbable entry of Japan into the ILO in 1919, when it did not allow union representation, and its refusal to join the Asian Infrastructure Investment Bank in 2016, even as the government increased spending for infrastructure in the region. The case research draws upon archival materials and testimony in the Japanese legislature for a range of perspectives on how foreign policy ties emerge as a prominent consideration in the discussion of IGO membership. Regional boundaries are the most frequent criteria for international organizations that form subgroups of states. Yet the political construction of regions occurs when geography becomes a discretionary criterion for membership in

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 33

regional organizations. Under the facade of common location, some regional organizations represent discriminatory clubs in disguise. Chapter 7 examines the role of foreign policy in determining membership in regional organizations relative to the importance of democracy, economic ties, and geographic location. Statistical analysis of 197 regional organizations shows evidence that security ties have as large an effect as distance and democracy in correlating with membership. Two case studies, of the EU and ASEAN, provide more detailed analysis of entry decisions. The chapter highlights exceptions to meritocratic review that occur even in the EU, which represents an ideal type of conditional entry based on democracy promotion—the 1963 French veto of British entry, mixed strategies in the Eastern enlargement, and resistance to Turkish membership. Brexit is discussed as an outlier case for theories of integration and for the argument of this book. While not a divorce over divergent security goals, foreign policy differences were one factor within the broader context of sovereignty concerns. In contrast to the EU, ASEAN never presumed to establish hard criteria for membership, and its enlargement occurred through gradual expansion across a set of governments that range widely in political and economic conditions. Formed to strengthen resistance to communist Vietnam, ASEAN would eventually embrace Vietnam as a member. In both the EU and ASEAN cases, foreign policy shifts were a critical antecedent to membership changes. Rather than constituting fixed standards for entry, both democracy and geography are selectively applied as conditions for membership in regional organizations alongside foreign policy criteria. Through the political process of selecting members in regional organizations, states shape the most fundamental parameters of regional identity. In a hard test for the argument about IGOs as discriminatory clubs, chapter 8 focuses on those international organizations that embrace the goal of including all states. For cooperation to provide a true public good, universal IGOs have open eligibility for accession. Yet here, too, membership terms set limits on universality. The community of states relies upon membership in IGOs to define sovereignty, and this means debates over membership take heightened significance for contested territories such as Palestine or Taiwan. This chapter explores how entry into universal organizations relies on positive relations between states and not simply on the capacity to participate in the cooperative endeavor or on legal standing through diplomatic recognition. Statistical analysis of entry into 88 universal IGOs shows that security ties correspond to higher probability of entry, which indicates that even

34 c h a p t e r 1

universal IGOs undertake discriminatory politics to favor allies. Four case studies examine membership for contested states. The longstanding controversy over membership in the United Nations for North and South Korea was only resolved with their simultaneous admission in 1991 at the end of the Cold War. Taiwan serves as the extreme example of a high-performing government that has been excluded on the basis of its sovereignty conflict with China rather than any concern about compliance. The case study examines how the United States and political leadership in Taiwan have influenced when and how the country joins international organizations. Taiwan’s struggle to participate in the WHO takes on greater significance amidst the pandemic. The movement to expand recognition for Palestine in IGOs shows divisions along political lines irrespective of the qualifications of Palestine to join any individual organization. Identity and rivalry also play a central role in debates over termination of membership, which is explored in the case study of South Africa’s expulsion from the UPU as a sanction against Apartheid. Variation across IGOs for these cases helps to illuminate the tradeoffs between efficiency for cooperation and reconciliation of sovereignty conflicts. The final chapter explores the implications of IGO member politics for our understanding of international institutions and cooperation. When institutions form discriminatory clubs, cooperation expands along familiar cleavages of security interests. Geopolitical rivalries also generate balancing in the sphere of multilateral governance. The conclusion focuses on three larger questions. First, to what extent do institutional ties reinforce or even aggravate geopolitical divisions? The chapter reflects on the role of membership selection to shift how we interpret studies about security cooperation. Second, does the practice of discriminatory membership promote international cooperation? On the one hand, unexpected joiners expand participation in cooperation, but on the other hand, they reduce the effectiveness of rules. This highlights the difficulty of accurately assessing the net gains for policy outcomes and the legitimacy of international organizations. Third, what are the dynamic trends for IGO membership patterns? Given an underlying demand to discriminate, expanding membership in IGOs will in turn generate new pressures for the proliferation of international organizations. States create new institutions as a tactic to continue the practice of discrimination and adapt to the problem of overexpansion. The answers to these questions open a broader agenda for research into how states use institutions to structure international society.

m e m b e r s h i p a n d i n t e r n at i o n a l c o o p e r at i o n 35

1.5 Conclusion The book considers IGO membership from the perspective of security interests along with institutional design and effectiveness. The politics of joining organizations touch closely on concerns about consolidating relations with other states as well as on functional demands for cooperation in complex issue areas. Drawing on insights from economic theories of club goods and sociology theories of networks and status, the book offers new insights into how states choose the groups they belong to and why these decisions have farreaching consequences. At one level, the security logic of IGO membership challenges institutionalist theories by introducing a different source of demand for institutions. Nevertheless, this opens the possibility for more impact by the institution on behavior because entry is not simply derivative to the policy interest in the issue area. States that join for other reasons related to foreign policy bring upon themselves pressure to reform policies. Entry triggers domestic mobilization of interest groups, reputation incentives, and socialization. Institutional theories explain how these mechanisms support cooperation. The role played by foreign policy at the time of entry increases the scope for unexpected reforms brought by the institution that are distinct from what the state would otherwise have done outside the institution. Furthermore, recognizing the foreign policy role of IGOs should moderate any assessment about institutional effects. When states form club-style IGOs that favor entry by friends, they lower the level of cooperation in the issue area relative to a meritocratic process with rigid conditions for reform as condition of entry. But it would be wrong to then conclude the institution is ineffective—states have chosen to prioritize other purposes related to status and economic statecraft over maximal gains from issue area cooperation. Evaluating outcomes based on policy reforms and gains observed within the issue area neglects the broader foreign policy benefits achieved by participating in the organization. At the same time, any pacifying effect attributed to IGO membership may arise from the earlier selection mechanism based on common geopolitical interests. Without attention to selection, researchers will miss larger effects of joining IGOs and may reach the wrong conclusion about the mechanism for the correlation between IGOs and peace. As a result, geopolitical discrimination is inefficient from the narrow perspective of cooperation on the regulated issue, but can generate benefits for IGO cooperation through increasing the likelihood that states enjoy positive

36 c h a p t e r 1

relations. The dysfunctional aspect of politicized entry into institutions is the risk of overexpansion. Logrolling to allow entry of friends before they have completed reforms contributes to broader expansion from what one state would choose as the optimal number of cooperation partners. Consequently, use of IGOs as a discriminatory club could yield mixed overall welfare effects. This book explores why club-style IGOs are prevalent and what determines who joins the club. The geopolitical logic of IGO membership accounts for both unexpected reformers and inefficient expansion of IGO membership beyond clubs. Supporting evidence can be found in the evolution of membership in organizations and in country-level decisions.

2 Flexibility by Design: Rules for Accession

clear boundary rules form the first design principle for cooperation. According to Ostrom (2000, p. 149), “using this principle enables participants to know who is in and who is out of a defined set of relationships and thus with whom to cooperate.” Only after setting boundaries can a group allocate rights and obligations. Sometimes the process to differentiate between in-group and out-group is straightforward, but how can states in international society reach such an understanding? In her work on local communities, Ostrom notes, “Group boundaries are frequently marked by well-understood criteria, like everyone who lives in a particular community or has joined a specific local cooperative. Membership may also be marked by symbolic boundaries and involve complex rituals and beliefs that help solidify individual beliefs about the trustworthiness of others.” To the extent that states form a small community with a long history of interaction, they establish regional associations or commodity cartels in a process not unlike that in villages. As collective actors, however, states also confront unique problems to sustain cooperation. Systemic competition for power prevents centralized governance, while repeated interactions necessitate policy coordination. States balance these contending forces through careful selection of partners for cooperation in international organizations. When organizing an institution to provide a public good, the goal is to maximize participation for mutual gain. But many policy problems are impure public goods—they are neither fully public with common benefits shared by all, nor are they private with zero-sum rivalry over consumption. Instead, excludable benefits and congestion characterize club goods where a smaller 37

38 c h a p t e r 2

group can share in the production and utilization of the good.1 From golf clubs to cartels, exclusion mechanisms support joint contribution while limiting congestion. Most international organizations are designed for the provision of club goods (e.g., Keohane and Nye, 2001; Greenhill and Lupu, 2017; Viola, 2020). They erect entry barriers and set approval processes to restrict membership. This book builds on this insight to portray how IGOs act as discriminatory clubs that select partners who have positive attributes on other dimensions distinct from criteria directly related to cooperation on the regulated policies. The process of joining international organizations often resembles entry into a social club that has the ability to deny entry based on informal decisions about who is the right type for the club. Members care about not only provision of the cooperative good but also the identity of other members of the club. Therefore, rather than screening for states with the highest capacity to share in the production of the cooperative good, states allow entry for those who share their values, regime type, and foreign policy, while excluding others. Flexible rules facilitate this process. Much attention in international relations has shown that hierarchy and discrimination can remain within an order based on sovereign states. Powerful states establish rules to reinforce their authority, and these interests shape exchanges within international institutions and exclude some from joining the order (Gilpin, 1981; Lake, 2009; Lascurettes, 2020; Viola, 2020). By varying the level of delegation or pooling of authority, states can set bounds on the authority ceded to the international organization (Hooghe, Lenz, and Marks, 2019). Multilateral norms of nondiscrimination, however, restrain efforts to distort the allocation of benefits within an institution. Much of the authority of international institutions derives from their ability to appear legitimate and exert normative pressure over state behavior (Hurd, 1999). States have tried to reconcile these tensions between hierarchy and legitimacy through their design of international institutions. (Martin, 1992; Koremenos, Lipson, and Snidal, 2001). Membership politics represents one of the foremost ways for states to embed discrimination within cooperation. Hurrell (2004, p. 41) writes, “Control over the membership norms of international society and the capacity to delegitimize certain sorts of players through the deployment of these norms represents a very important category of power.” A critical decision to shape 1. See Buchanan (1965) and Cornes and Sandler (1996).

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 39

the organization, membership has not been delegated to bureaucrats within the organization.2 The informal nature of discrimination over membership preserves the principle of multilateral cooperation that is the basis for legitimacy. After limiting who can join, states can more readily support multilateral norms within the organization. This chapter explains the variation in the design of membership provisions for international organizations. My goal is to evaluate the hypothesis that states will design IGO accession rules to provide discretion over selection. I focus on two related questions. First, under what conditions do states pursue open multilateralism? To the extent that cooperation represents a public good, one would expect the prevailing form of membership to be open membership without restrictions. The frequency of selective IGOs suggests instead that many of the cooperation problems in international affairs represent club goods. This leads to my second question: when states restrict membership, how do they design the exclusion mechanism? Precise conditions for entry paired with provisions for suspension and expulsion could support deepening cooperation with high compliance. A third-party review committee could assess suitability for membership based on inspection of implementing legislation. But this kind of meritocratic screening is surprisingly rare. More often, the provisions for membership are vague about the terms of entry and rely on votes of approval. I argue that this discretion allows members the flexibility to exclude states for reasons outside of the formal scope of the institution. Just as social clubs do not write down the rules that would differentiate on the basis of race or ethnicity, IGOs do not formalize the criteria for entry. The vague terms and room for exclusion at time of entry represent a flexibility mechanism in the institutional design. Ad hoc criteria facilitate statecraft that uses IGO membership as carrot and stick to advance purposes beyond cooperation on the issue at hand. Empirical analysis in subsequent chapters of the book demonstrates how states have used this flexibility to inject geopolitics into economic organizations like the OECD and WTO. Turkey found early entry into the OECD possible as a NATO ally. Iran’s application to join the WTO in 1996 was blocked by the United States under the guise of its needing more time to review Iran’s trade policies, even while Afghanistan’s application in 2004 reached its conclusion despite major obstacles to effective 2. In their analysis of decision procedures for international organizations, Hooghe et al. (2017, p. 130) find the least delegation for accession and suspension relative to policy-making and budget areas.

40

chapter 2

compliance. One can imagine that different outcomes may have resulted from a meritocratic selection by a review commission assessing performance criteria. Yet even as legalization trends are characteristic of enforcement in the WTO and many other IGOs, few have delegated the decision over screening members to third-party assessment. This chapter develops an argument about the design of membership provisions in section 2.1. It explains why membership criteria are deliberately vague and prone to discrimination. Section 2.3 introduces data from over 300 IGO charters on the design of accession rules. In section 2.4, I discuss the patterns for the design of IGO membership rules and consider dynamic questions related to the evolution of accession processes and use of informal practices.

2.1 IGO Accession as Club Membership What we know about IGO membership has been developed within the context of studying cooperation problems in international relations. Theories address the question of why states choose to establish institutions and the conditions that influence compliance with rules. Empirical research provides extensive analysis of ratification patterns and the impact of treaties on behavior. These are related questions, but distinct from membership. On the one hand, membership politics are a subset of cooperation problems that arise around the decision to accept an agreement. The key variables of power and information asymmetry that underlie existing theories about cooperation also shape membership politics. On the other hand, membership politics represent a unique dynamic shaped by entering a formal organization with a particular group of states. Relational variables become important because the decision is not only about whether to cooperate on an issue but also about who should work together. International organizations allow states to pool their authority through collective decision-making. We can observe wide variation in the degree to which members retain control, but an entrant into an organization accepts common participation in governance activities (Stone, 2011; Hooghe et al., 2017). Beyond the commitment to uphold the signed agreement, membership includes a promise of forward-looking engagement to enforce rules and deliberate about new policies. In this sense, the comparison of treaty ratification versus accession can be seen as the difference between firms making a contract for a single project and those forming a joint venture. Both agreements would commit actors to terms for pricing and activities, but the latter also allocates

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 41

managerial authority. This dimension of shared authority makes selection of partners one of the most important parameters of cooperation within an international organization. The design of accession rules by the states that found the organization emerges out of their own choice of scope for cooperation with other states. Once enshrined in a founding charter, the membership provisions lock in this selection criteria—or defer the evaluation of what it takes to qualify for membership by not specifying selection criteria. The conventional view in international relations theory addresses two types of IGO—universal organizations that aim to provide global public goods and smaller organizations representing a subset of states that share preferences for cooperation on the given issue. The presumption is that the former include all states while the latter engage in rigorous screening for expected compliance. Neither fits reality. Most IGOs fall into a form of club IGO, where membership is neither universal nor based on rigorous screening. These are discriminatory clubs; they evaluate members in terms of their ability to contribute to cooperation within the issue area and their attributes outside of the issue area (Cornes and Sandler, 1996, p. 385). The discriminatory club model introduces a relational component to the membership equation. The value of association with another state makes others support its membership independently of its expected contribution to collective action for the cooperative outcome. Income, race, or other features could enhance the value of association. The point is that states may want to design a selection process such that they can choose their own criteria for membership. When the criteria are unrelated to the mandate of the organization, selection will become discriminatory. Instead of evaluating expected performance ability, actors inject their own preferences for whom to associate with through interaction in the club. In a period of heightened attention to racial justice, we are increasingly aware of the ripple effect of racial discrimination in causing harm to individuals and society. These patterns of bias become embedded in institutions that perpetuate the gap in opportunity. On top of the moral objections to discrimination, one can also condemn its negative effect on welfare. Becker (1971) pointed out the economic inefficiency of racial discrimination, since hiring based on attributes unrelated to productivity would lower output. Applying the economic method to analyzing society, such behavior is rational for the individuals who practice it as long as one widens the assumption of preferences to include what Becker termed a “discrimination coefficient.” This takes into account the possibility of prejudice, making race, gender, or other

42

chapter 2

attributes influence employers and customers outside of the characteristics of goods and services provided. The framework encompasses discrimination to favor an individual based on non-market attributes. In international affairs, the use of alternative criteria for selection outside of performance could also constitute a discrimination coefficient. As with employment discrimination, it would reduce the effectiveness of regime performance in the issue area but remain rational. The alternative criterion may achieve other goals. There can also be cases where discrimination advances moral causes, such as affirmative action policies, or excludes actors as a punishment. As we will see in a later chapter, one case of IGO discrimination arose over the expulsion of South Africa from the UPU as a sanction for the injustice of Apartheid. It is not my purpose to judge the morality of states when they diverge from a rigid performance evaluation. The analysis here focuses on why and how states discriminate and the features that make countries attractive for joint membership. When states design international institutions, membership conditions range from inclusive to exclusive participation, and they also vary in the rigor of their requirements for entry. Four models characterize the types of membership design: meritocratic IGOs establish strict conditionality for compliance in order to prevent free riding (Downs, Rocke, and Barsoom, 1998), hierarchical IGOs allow variable commitments and authority based on capacity (Gilligan, 2004), club-style entry uses attraction and selection with flexibility for response to changing circumstances, and universal organizations welcome all who are acknowledged as members of a community. The first two conditionality models focus on performance within a single-issue dimension, whereas the latter offer more discretion, with flexibility for issue linkages to geopolitics that raise and lower the bar. Institutional theory and realist skeptics both presume the prevalence of conditionality or universal organizations, but a surprisingly large and diverse range of organizations follow the club model. Figure 2.1 displays four ideal types based on the variation in participation mandate and conditionality terms. The participation mandate sets parameters for whether all states are in principle eligible for membership and establishes the expectation of whether the IGO aims to achieve narrow or broad membership. The conditionality terms establish criteria for the accession process, which can be precise benchmarks identified as objective measures or may be left to the discretion of members. Precise conditions effectively serve to limit participation, but they could simply codify a broad set of eligibility criteria that any state could meet. Furthermore, an organization could be selective,

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 43

Conditionality terms

precise

vague

Hierarchical (IMF)

Meritocratic (EU)

Universal (WHO)

Club (OECD)

Inclusive

Selective

Participation mandate

figure 2.1. Four models of IGO accession process

without offering details on how it restricts membership, or it could be inclusive, while requiring steps necessary to gain entry. If one takes the analogy of the university admissions process, conditionality terms are more vague for Ivy League schools in the United States, which rely on a wide range of evaluation criteria, than for a national university in Japan, which relies on test scores, even though both are highly exclusive in limiting admissions. In the realm of extracurricular activities, a ballet troupe and a tap dance club may both have similar policies for club tryouts (precise entry terms), while the former is more exclusive in limiting membership to a small group, and the latter takes an inclusive approach, allowing participation by any who show up. The categories hold heuristic value to contrast alternative models. Yet even the IGOs given as examples of meritocratic or universal IGOs fail in practice to uphold such principles. Case studies in this book will reveal evidence of discrimination over membership within both the WHO and the EU. A universal IGO claims to be open to all states. Global public goods issues such as health and the environment are leading areas for the univeral IGO structure. Climate change negotiations face the challenge that nonparticipation by large developing countries will prevent others from achieving their goal to stop the rise of global temperatures. The Covid-19 pandemic shows how diseases in any country could affect everyone, which makes it important

44

chapter 2

for all states to share information with the WHO. Unable to exclude benefits or harm, states will favor open eligibility for a universal organization in the effort to support global public good provision. The legitimacy of these IGOs depends in part on the perception of their universal principles and fair procedures (Lenz and Viola, 2017; Tallberg, 2019). Nevertheless, the use of membership as an organizing concept implies distinguishing between insiders and outsiders. Welcoming all states still requires deciding what counts as a state. Sovereignty is a multifaceted concept without a single definition (Krasner, 1999). States may harbor intense disagreements over this question, such as those witnessed over applications by Palestine and Taiwan to the United Nations. The affirmative outcome of membership decisions of the universal IGOs offers recognition of sovereignty. Even universalism has its limits, which are all the more intense because the vague mandate of eligibility leaves open to members’ discretion the decision to define what counts as statehood. Hierarchical organizations are inclusive in mandate and limit membership benefits through conditions that determine influence over decision-making.3 The IMF requires states to contribute a subscription share to support available funds. The fee-based entry of the IMF goes beyond mere requirement to pay dues. Each member entering the IMF is assigned a quota based on its size in the global economy that determines the financial resources it shall commit and its voting share.4 This formula offers a way for the institution to remain open, while weighting influence toward those with the greater share of duties. An alternative fee-based formula would rely on utilization rates, similar to a toll road. Remarkably few IGOs apply this particular form of variable membership entry requirements. One example would be the Advisory Center on WTO Law, an organization that charges members dues to support the organization and hourly fees for use of legal staff to support the government in WTO dispute settlement proceedings. While utilization fees such as those of the ACWL are unusual, budget dues are a common requirement of membership, with nonpayment constituting reason for suspension of membership privileges. Budgetary dues are typically allocated in proportion to income, which is, if anything, the reverse of utilization rate, given that governments of lowincome states are more likely to rely on IGO services. The regular assessment 3. I thank Stephanie Rickard for clarification on this point. This category encompasses the heirarchical multilateralism and exclusive multilateralism described by Viola (2020, p. 85) as strategies by which states expand membership but limit access within the institution. 4. Although even the allocation of quota rights and decision authority is not as objective as it may first appear, and has been the source of contestation (Stone, 2011; Lipscy, 2015).

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 45

of dues is characteristic of all four entry types, and it is only when contributions generate differential categories of membership benefits that I consider it a form of selection. The upper right quadrant of the figure shows the meritocratic IGO that is both exclusive and sets clear conditions for membership. The closest approximation can be found in the EU enlargement process following the principles established at the 1993 Copenhagen meeting of the European Council. The participation mandate restricts eligibility by both region and regime type. The accession negotiations involve European Commission officials undertaking a detailed review of compliance with each of the 35 chapters of the acquis that represents the body of EU legislation. This type of membership design presents many advantages for cooperation—an exclusive mandate lowers collective action problems by focusing on a smaller number of states, based on common traits. Precise entry conditions make explicit the practice of screening for compliance at the start of membership. The process itself becomes more bureaucratic with delegation to IGO officials. To maximize the depth of cooperation achieved within an issue area, one would expect states to pursue this form of membership design. It is therefore surprising to observe that very few IGOs adopt this design in their membership provisions. Even the EU deviates when it holds some countries in the waiting room after having started accession talks, or jumps forward Ukraine as an eligible country in a gesture of solidarity during war after years of insisting its policies fell short of EU standards. There are elements of discrimination that seep into the most meritocratic process. The final category includes club organizations that are exclusive but set only vague conditions for entry. Social organizations such as a fraternity clubs or elite groups like the Council on Foreign Relations do not specify the qualifications for membership. This does not mean that there is not a review process that can involve hazing rituals and reference letters to assess applicants. The OECD stands out as an example of an international organization that follows club-style entry provisions, with the OECD charter Article 16 stating: “The Council may decide to invite any Government prepared to assume the obligations of membership to accede to this Convention. Such decisions shall be unanimous.” In the case of the OECD, countries’ policies are carefully examined. Nevertheless, decisions about whom to invite to enter accession negotiations reflect a complex calculation about whether a state is “like-minded” and a “significant player,” and not simply the conformity of policies with the many OECD conventions. Other organizations dispense entirely

46

chapter 2

with any review process to assess compliance with terms of the agreement; the deposit of a treaty protocol means commitment in principle but not proof of adherence. For club-entry IGOs, the organization charter provisions on accession typically require acceptance of the organization principles and approval of members, without offering any restrictions based on review of policies. The addition of membership approval allows for the possibility that a state could meet all terms of the organization agreements, but still be denied membership for other reasons. The majority of IGOs fall within this range of moderate selectivity and vague conditionality. In short, the process to enter most IGOs is much closer to getting into a social club than trying out for a sports team. The advantages of club-style accession processes are twofold. First, members maximize their discretion over enlargement. Without any need to write down new rules, states can extend preferences to former colonies or allies across a range of issue areas through selective membership. Discretion allows states to change the criteria over time, such as shifting to consider democracy or other conditions that may not have been important at formation of the regime. Using their discretion over entry criteria, members can adjust across multiple criteria rather than holding to a single rigid standard. A state that meets one standard could fail on another. Second, members retain their bargaining advantage through not being transparent about the accession process. Uncertainty over what must be done to join the organization can be used to extract more concessions during accession negotiations (Schneider and Urpelainen, 2012). There are also advantages to the applicants. First, the approval or rejection of an application carries information for the applicant about their social position in the international system because the process allows room for favoritism and discrimination. Amidst the uncertainty of anarchic international relations, institutions establish affiliations through mutual cooperation for governance. Common institutions form the basis of an international society and can become a tool in the maintenance of hierarchy (Bull, 1977; Lake, 2009; Viola, 2020). Joining organizations forms one kind of signal that states have a working relationship based on mutual recognition. Conversely, rejection implies that other states do not view the candidate as a valuable partner. For states seeking information about their threat environment and the hospitality of the international system, such signals of approval or rejection carry weight beyond the business of the organization. Second, entry will signal status more effectively when membership is not simplyarubberstampingofobjectiveconditions.Enteringintoanorganization

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 47

is equivalent to choosing a social category for belonging to a group. Not only are states accepting specific commitments, they are also choosing to associate with a particular group of states. Indeed, the social context lends more power to the norms of compliance, as individuals receive utility from conforming to group expectations (Akerlof and Kranton, 2010). This same process is posited to operate at the level of states as they interact within an organization (Johnston, 2001). The two-sided nature of membership reveals both that the state seeks this deeper association and that other states accept its entry into their community. As a choice of partners with whom to associate, IGO membership shapes the views of others about foreign policy, economic risk, and regime transition. Most directly, the IGO membership portfolio of a state identifies its foreign policy orientation. New choices can signal shifts in foreign policy direction. Kinne (2013a, p. 664) argues that IGO-based signaling reduces conflict between two states as they perceive entry as a meaningful step for cooperation even if the organization has nothing to do with security policies. The states of Eastern Europe and former Soviet republics had to navigate their transition to a new foreign policy identity closer to the West, and expanding their IGO memberships to include more European-centered organizations was a critical part of this process—for a state like Georgia, these steps were “deliberate Georgian efforts to signal an interest in cooperation with the West.” Boehmer, Gartzke, and Nordstrom (2004) find that membership in structured IGOs with formal rules and procedures reduces the probability of conflict onset. They posit information transmission improves the bargaining environment so that states can avoid conflict. The most pertinent information conveyed by IGO membership is the voluntary decision by a group of states to establish closer relations through their participation in a formal organization. Association with other states also conveys information to domestic and international audiences about the direction of regime transition—just as democratizing states may signal their intentions through joining IGOs with strong democracies, non-democracies participate more in IGOs with other non-democracies (Mansfield and Pevehouse, 2006; Libman and Obydenkova, 2013).5 Finally, joint membership conveys information to markets. Gray (2013) finds that peer effects from association in an IGO have substantial impact on how financial markets perceive the risk of a country, beyond the 5. Not all states are allowed to join—Poast and Urpelainen (2013) highlight that democratizing states often have to form IGOs when unable to join existing ones.

48

chapter 2

objective policy measures of risk characteristics. Investors looking for cues are quick to classify states into groups, and membership forms one such benchmark (Brooks, Cunha, and Mosley, 2015). The aspirational desire for peer effects and self-binding to the norms of a particular group of states contribute to demand for entry into club IGOs. In contrast, universal or meritocratic processes carry less information about a state’s social position. IGO membership forms the tool for allocation of status within international society. The dual basis for membership based on performance and association value makes it a more meaningful status indicator than if membership were only based on one vector. Performance alone does not require separate validation. The relational component is based on prior ties well known to both sides. Clearing both thresholds marks a high level of approval for the position of a state relative to others who are not members. Performance and relational ties evaluated at the group level produce membership outcomes that contribute to status at the international level.

2.2 International Society and Ending IGO Membership Exit and expulsion represent the other side of membership politics. The right to exit reflects the voluntary nature of cooperation in anarchy. Just as few states are coerced into joining an organization, as sovereign actors they retain the right to leave. Similar processes to those that shape entry into an organization also arise for ending membership—losing interest in the issues at stake within the organization or shifting away from shared values and geopolitical alignment. As preferences diverge, a state will judge the gains of membership are no longer worth the cost of compliance or budget fees, and either exit or lapse into noncompliance. Several theories about international institutions emphasize the willingness of states to exit as a critical constraint on cooperation (Stone, 2011; Johns, 2015). In the case when deviation leads to noncompliance rather than exit, current members face the question of whether to suspend or expel the state to eliminate free-riding and to uphold the cohesion of the community. Expulsion is more difficult as a collective decision of members to reject continued cooperation with a state. This section examines how the club logic of IGOs appears in the domain of exit and expulsion. Some stickiness of membership is to be expected—path dependence and sunk costs endow institutions with an inertia that helps them outlast the configuration of interests at their formation (Krasner, 1983; Jupille, Mattli, and Snidal, 2013). Indeed, some organizations remain in operation long after they

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 49

cease to deliver major gains from cooperation—Gray (2018) shows that these zombie organizations account for a substantial share of international organizations. Even in a context of sticky institutions, however, one would expect some attrition. On the demand side, shifting preferences and power balances can diminish the gains from participation for members. On the supply side, members should exclude those that routinely violate rules. How do the rules over membership address this potential for exit and expulsion? Agreeing on the terms over exit as part of the IGO charter assures other states that they will receive notification of any change. Cooperation occurs through the mutual adjustment of policies, so when one side ends cooperation, others will want to adjust. A universal organization that asks little of states to enter may still request prior notification of exit. Most recently, this was seen in the case of the U.S. declaration that it would leave the WHO. President Trump first announced his intentions to leave in May, and then in July 2020 the government issued the formal notification to the WHO. Rules for a one-year transition period meant that the U.S. exit would have taken effect a year later, in July 2021. Given the reversal of the decision by the Biden administration, however, the United States never actually exited the organization. In the case of the United Nations Educational, Scientific and Cultural Organization (UNESCO), the United States has exited the organization twice; it withdrew in 1984, criticizing poor administration as well as pro-Moscow and anti-Israel leanings on policy, only to reenter in 2002 when building international support for the war on terror. The Trump administration announced plans to exit in October 2017, with formal membership ending in December 2018. This second exit episode was the culmination of opposition to UNESCO membership for Palestine, and has yet to be reversed (see chapter 8). The provisions on exit are discretionary for the state choosing to leave without a vote by IGO members. The threat to end cooperation by expulsion marks a stronger form of conditionality. In bargaining models, a grim trigger strategy describes the termination of all cooperation as a response to defection by another actor. For membership organizations, this would take the form of expulsion. Looking at local cooperation based on gradation of sanctions for noncompliance, Ostrom (2000, p. 151) notes that “the capability to escalate sanctions enables such a regime to warn members that if they do not conform they will have to pay ever-higher sanctions and may eventually be forced to leave the community.” The possibility looms over interactions in a group even if it is not frequently used or written in the treaty. Nevertheless, to the extent states seek credible

50 c h a p t e r 2

commitments that signal future behavior, one would expect formalization of terms for expulsion. Expulsion could complement precise entry conditions or serve as a substitute. On the one hand, an organization with precise entry conditions and limited eligibility could gain leverage to enforce those provisions through parallel terms for ending membership when violated. On the other hand, an organization with open entry rules could review member policies and expel low performers to uphold compliance. From a contractual perspective, expulsion terms should feature prominently in IGO charters, and it is surprising to note their relative infrequency. States largely treat the question of ending membership as a separate matter from compliance. This is reflected in both the design of IGO rules and the behavior of states. In a pattern that more closely resembles citizenship than a contractual relationship, few choose to exit and even fewer find their violations of rules and principles sanctioned by expulsion. Despite the low threshold for exit, it rarely occurs. In a study of IGO treaty withdrawals from 1945 to 2014, Borzyskowski and Vabulas (2019b, p. 339) document 200 instances, which represent only 0.04 percent of IGO-member-years. The most frequent exiters—the United States and Canada—have only withdrawn ten and nine times respectively. Even more stark is the absence of expulsion terms in agreements. Surprisingly few IGOs include specific provisions for expelling states from membership, and when present, such provisions are rarely acted upon. This is readily apparent for some human rights organizations, where the inclusion of states with terrible records for abuses attracts scorn, but supports the belief in universal norms. When exit and expulsion do occur in international organizations, it rarely is about compliance. The most prominent cases of ending membership seem to be less about contract enforcement than a divorce in the relationship, such as the Brexit departure from the EU. An extreme cases, the expulsion of South Africa from the UPU, represented anti-Apartheid foreign policy sanctions rather than enforcement of postal rules.6 In part, the dearth of expulsion for noncompliance reflects the use of alternative punishments. Members exercise specific reciprocity or peer pressure to influence behavior from within an organizational context rather than expelling 6. As part of the international pressure applied on South Africa to end Apartheid, the members of the Universal Postal Union voted to expel it from the organization in 1964, despite objections from some who noted that expulsion was supposed to be limited to states shown to violate UPU regulations (“Universal Postal Union,” 1966, p. 834). This case is examined in chapter 8.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 51

noncompliant states. Partial responses occur in the kind of graduated sanctions expected by Ostrom (2000). One such measure is the temporary suspension of voting rights, such as the OAS’s actions against Cuba. Yet even this action requires a strong consensus among members; in their comparison of IGO charter rules, Hooghe et al. (2017, p. 125) find that IGOs have the least amount of pooling for decision-making related to suspension. Another response to noncompliance is the withdrawal of benefits proportional to the degree of violation, such as in WTO dispute settlement. Differential levels of benefits allocated among members can reward some and punish others (Stone, 2011; Vreeland and Dreher, 2014). As Europe debates democratic backsliding in Poland and Hungary, conditional subsidies have emerged as one leverage point. Finally, the continuity of state membership masks other forms of dynamism in who cooperates that arise by changing recognition of a government. The dissolution of the USSR and its replacement by Russia occurred without a new accession process. According to Landau-Wells (2018, p. 103), civil wars often bring a change of recognition that precedes victory on the battlefield. Others may refuse to acknowledge a ruling government, as in the controversy over representation of Afghanistan and Myanmar in the United Nations that prevents the ruling government from participation in meetings.7 This flexibility supports ongoing membership, although nonrecognition may form de facto expulsion. These lower-level enforcement actions play an important role as states monitor and sanction noncompliance. But more generally, the restraint about using membership as a tool to punish noncompliance reflects the nature of organizations as institutions that build on community. When ties go deeper than policy coordination, they are not used for enforcement. From the perspective of IGOs as discriminatory clubs, ending membership is also about the relationship among countries. The frequent analogy of Brexit as a divorce highlights the notion that positive or negative relations motivate exit. In order to become members in a discriminatory club, actors value joint association as well as the collective effort for provision of public goods. Factors that diminish the value of association can strain perceptions about continued cooperation in the IGO. For individuals, income, race, or religion may form indicators of shared status. A fall from fortune or change in social norms toward in-group identification may shift individuals to lower attractiveness as fellow members in a club. 7. Gladstone, Rick. “U.N. Seats Denied, for Now, to Afghanistan’s Taliban and Myanmar’s Junta,” The New York Times, 1 December 2021.

52 c h a p t e r 2

For governments, peer status as like-minded countries hinges upon ideology, regime, or geopolitical alignment. When a regime has a revolution or breaks alliance ties, this changes the calculus for assessing the value of association with that country. A turn inward toward isolationism corresponds to retreat from international organizations. When others come to view a state as a pariah unfit for peer association, this can build momentum for the collective decision to expel it from membership. The relational value of membership in a discriminatory club offsets the attention to policy. On the one hand, this could reduce the need to include expulsion terms for noncompliance in the charter. When other states receive reputational spillover from association with other members, it weakens the link between compliance and membership. They can continue to cooperate with a noncompliant state that remains in good standing in terms of political relationships just as they may choose to terminate cooperation with a fully compliant state because it has violated community norms. On the other hand, the discriminatory club may expel members when association with a member takes on negative value. Vague entry rules facilitate formation of a club of like-minded actors along multiple dimensions, including some unrelated to cooperation, and vague rules on expulsion also support the termination of cooperation without due cause so long as the membership votes one out of the group. Omission of rules on expulsion leaves such actions at the complete discretion of members. Charters with vague or missing terms for expulsion are consistent with a club model of IGO membership. The threshold of conflict that could lead to ending IGO membership is quite high as a result of collective decision-making processes that favor the status quo. With the need for a supermajority of members to vote for suspension or expulsion, broader ties across the entire membership matter. A state may witness criticism and deterioration of relations with some states but not yet trigger opposition from sufficient states to warrant expulsion. As each state fears itself being the next target for such votes, they may not readily join a coalition to expel another. In some institutional contexts, smaller states can exert a surprising degree of influence through their blocking power— multilateral institutions are inherently prone to this problem of multiple veto players that block policy changes (Tsebelis, 2002). This is widely evident within the European Union. The nature of EU decision-making rules has shifted the relative influence of states such that a country that may not have power in the traditional sense could nevertheless shape decisions through its outlier preferences on policy (Garrett and Tsebelis, 1996; Scharpf, 2006;

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 53

Schneider, 2009). Enlargement adds to the challenge of decision-making by creating even greater preference heterogeneity. Yet when states become dissatisfied with open noncompliance, other members can block reforms, including those related to membership. In the most sensitive issues of suspension over democratic backsliding, veto players have largely protected Poland and Hungary. One may only appear to be a promising cooperation partner to join, while a state behavior must be extremely negative to draw calls for expulsion. The low incidence of expulsion carries implications for membership politics. In effect, states have created an obsolescing bargain, whereby once entry is granted, other members have less influence over behavior than they did at the time of accession.8 This places greater importance on the need to carefully screen entry into the organization. States establish formal membership provisions, which they can modify with additional practices that will condition outcomes. The particular historical experience within a region and examples from other organizations can lead to different combinations of institutional design rather than a singular template (Alter, 2014; Jetschke, 2017). Other studies document how enforcement provisions vary widely across institutions, and we see similar variety in the design of membership rules. Evaluating the range of design choices will help to test the prevalence of the club model. Attention to rules for entry and exit offers insight into how states do and do not condition membership.

2.3 Data on IGO Accession Rules This section uses descriptive statistics about the rules of accession to evaluate the first hypothesis underlying my theory about IGOs as discriminatory clubs: states will design IGO accession rules to provide discretion over selection. Further, I shed light on the two motivating questions of the chapter: when are IGOs universal, and how do restrictive IGOs screen for membership? Following the typology of four models of the IGO accession process (Figure 2.1), I categorize the means of exclusion mechanisms in terms of participation and conditionality. As a preview, the evidence presents an overall picture of discretionary rules that maximize the flexibility for members to choose how they screen members and documents the prevalence of the club model. 8. See Carnegie (2015), who models membership in international institutions as a constraint on the exercise of coercive diplomacy.

54 c h a p t e r 2

I also compare the pattern of accession rules for IGOs in different issue areas. The logic of discrimination in my theory arises across issue areas, but much of the literature on institutional design suggests that states adapt provisions to guide behavior appropriate for the kind of cooperation problem of an issue area (e.g. Martin, 1992; Koremenos, Lipson, and Snidal, 2001; Hooghe and Marks, 2015). In the environmental policy issue area, the cooperation problem is more likely to resemble provision of a public good where exclusion is not feasible. In the social issue area the distributional stakes may be lower such that exclusion would be unnecessary. This could be a limiting scope condition for IGOs to form a discriminatory club. The expectation from this literature holds that policies to regulate the environment and social issues would have a higher frequency of universal IGOs relative to economic and security issues. Examining whether discretionary rules and selective membership occur across issue areas is important for understanding the generalizability of my argument.

2.3.1

Founding Charter Documents

In order to research the design of institutional rules, I restrict my inquiry to those IGOs for which a formal document specifies the rules governing the institution. These documents are referred to by different names, including constitution, treaty, convention, agreement, statute, and charter. I will refer to them collectively as “charter,” which reflects that the text describes the basic rules and principles for the organization irrespective of its legal status. My analysis focuses on the design of membership provisions as written in the charter documents. Informal IGOs that lack a public record of a charter fall outside the scope of the analysis presented here. The population of IGOs for analysis draws upon the Correlates of War (COW) International Organizations Dataset version 3.0 (Pevehouse, Nordstrom, and Warnke, 2004). Starting from the 534 IGOs listed in the COW data for the period from 1815 to 2014, I coded accession provisions for 322 IGOs for which I could locate a charter document establishing governance rules of the IGO. The search for IGO charter included identification in the UN Treaty Series, the IGO website when available, and the volumes on International Governmental Organizations by Peaslee that are referenced as a source for the IGO database (Peaslee and Peaslee Xydis, 1979).9 The oldest available charter 9. I thank Raymond Hicks for this extensive search to assemble the complete charters for analysis. I have omitted IGOs that were entirely subsidiary to another IGO, such as the

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 55 300

Number of IGOs

250 200 150 100 50 0 Economic

Social

Environmental

Security

figure 2.2. IGOs by issue area: The figure shows the number of IGOs by the issue areas that represent part of the core mission. Note that IGOs may be classified under multiple issue areas.

forms the basis of coding membership provisions. My question of interest is about the institutional design decision of founders. Moreover, research into the changes of charter provisions over time for these organizations suggests that states rarely modify the membership provisions.10 The informal practices of how states evaluate candidates shift more than the legal constraints. The data describes the issue area of the IGO based on the aims expressed in the preamble of the charter and information from the Yearbook of International Organizations regarding the fundamental aims and subject area for all organizations.11 Some organizations with a broad mandate will address several issue areas, while others are more narrowly focused. Figure 2.2 presents the Inter-American Defense Board that operates within the Organization of American States. In this data, IGOs that have a substantial change of legal status are treated as new IGOs, i.e., the GATT and WTO appear as two IGOs. Two additional organizations that fit the criteria of an IGO according to COW guidelines were added: Eurasian Development Bank and the Global Green Growth Institute. 10. In their detailed history of 76 organizations for the period 1950–2010, Hooghe et al. (2017) found only 14 that changed provisions for accession (variable ACC17), with all but two of those changes simply moving from a status of “no documentation” to coding that there is no requirement for member ratification. The substantively meaningful change was the EU addition of a requirement that members must ratify an agreement for the accession of new members. 11. The Yearbook of International Organizations is a compendium of information on over 66,000 international organizations produced by the Union of International Associations. Coding relies on two categories of information: the goals of each organization (“aims”) and

56 c h a p t e r 2

distribution of the IGOs across four issue areas. Looking at the categories of economic, security, environmental, and social issues for the 322 IGOs in the sample, 80 percent refer to economic topics as a core aim. This in part reflects the broad definition of the category, which includes not only the WTO but also the UPU. Furthermore, there is a tendency for IGOs to link their purpose with economic development—the Global Environment Fund is classified as both an environmental IGO and an economic IGO. Only 11 percent of the IGOs fall within the category of security issues. Many alliances do not take the form of an international organization—NATO enters the sample while the U.S.-Japan security alliance does not. Nineteen percent of the IGOs incorporate environmental goals as a priority, with many holding this as their core purpose, such as the International Whaling Commission (IWC), and others, such as the African Union (AU), that uphold environmental sustainability as one of several goals. Social issues are highlighted for IGOs that may serve general purposes of regional integration, such as ASEAN, or for IGOs where they constitute the central focus, such as UNESCO and the International Organization for Migration. The overall size of the IGOs varies from small organizations, with a few members, to the United Nations, with 193 members. This range is apparent across issue areas. When facing the trade-offs between larger or smaller membership size, it is not evident that the nature of the issue plays a determinative role. The median organizational size is 20 members, which is similar to that of economic, social, and security organizations.12 Environmental organizations have a lower median size of nine members, which reflects the regional focus of several that address a specific river basin or fisheries area. Figure 2.3 presents the data for the total number of members in IGOs disaggregated into the issue areas.

2.3.2

Participation Mandate

There are three ways in which IGO design of the charter can restrict membership. First, the charter can establish a broad category of eligibility rules. These the issue area listed for the organization (“subject”). Text scraping software parses these descriptions for keywords. For example, keywords such as “commerce,” “development,” or “finance,” which indicate a focus on economic activity, would categorize the IGO to fall within an economic issue area. This text coding is supplemented by the descriptive coding of the IGO charter. 12. This focuses on the largest size achieved by the organization, whereas looking at the average size of an organization over its existence would show a smaller figure of 15 members as the modal size of organizations.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 57

Membership size (maximum)

200

150

100

50

0 Economic

Social

Environmental

Security

figure 2.3. IGO size by issue area: The figure shows the number of members in each IGO at the maximum size observed during its existence. Note that IGOs may be classified under multiple issue areas.

serve ex ante to limit the set of countries that would seek membership. Second, negotiable terms of membership allow for evaluation of policies as well as demands for additional reforms to make it easier or harder for a country to join the organization. Powerful states are likely to favor the use of negotiable terms that allow them to use bargaining leverage and informal influence to shape who can join and to determine the price of entry. Third, a requirement for member approval in the form of a vote ensures support for new members. Rules vary in terms of the size of majority vote required, and some call for unanimity. The requirement for an invitation to apply adds another layer of member approval. Negotiations and approval requirements offer members more discretion than do eligibility rules. IGOs may use a combination of these steps to restrict participation. Universal IGOs are open on all three dimensions of participation. For example, the WHO in Article 3 of its 1948 Constitution of the World Health Organization declares that “membership in the Organization shall be open to all States.” Even the IGOs with open eligibility often contain some restriction, such as the WHO requirement that the Health Assembly vote by majority to approve membership in the case of states that are not already members of the United Nations (UN members may simply deposit the signed agreement without vote approval).13 In the sample of 322 IGOs, only 40 fall in 13. The significance of restricting entry for recognized states is further developed in chapter 8. An important exception is actors that are not UN members. These often must receive approval by a majority, or in some cases a supermajority, of members.

58 c h a p t e r 2

the category of universal organizations, including those such as the International Telecommunication Union (ITU), UPU, International Labour Organization (ILO), and WHO. The vast majority of IGOs take measures to limit participation. Restricted eligibility. Eligibility rules form the most frequent restriction. Sixty percent of organizations have some regional focus within their mandate, although not all restrict membership to those within a geographic region.14 When regional IGOs restrict membership to the region, they often do not define the scope of the region in objective terms. For example, ASEAN declares that the “Association is open for participation to all States in the South East Asian Region” without specifying the borders of the region. Chapter 7 will analyze how discretionary rules allow geopolitics to play a role alongside geography when selecting members. The next most common form of exclusive eligibility is a charter restriction that refers to membership in another IGO—69 of the IGOs refer to another IGO such as the UN or a regional organization.15 A small group of 23 organizations such as OPEC regulate prices and related policies for commodity trade and restrict membership to those that are producers of the commodity. Yet even among the commodity organizations, a few, such as the IWC, lack such a restriction. It is equally important to note what kinds of exclusive eligibility restrictions are not present in IGO charters. Rules generally have not mandated a particular political regime as a condition for entry. The EU stands out for its requirement that member state governments be founded on the principles of democracy (Article F of the Treaty on European Union), but most organizations do not specify such a requirement, even when the members consist almost entirely of democratic regimes. For example, as discussed in chapter 5, the OECD has come to treat democracy as a priority in review of accession applicants, but non-democracies were among the founding members, and the formal charter provisions for membership do not specify any restriction by 14. Of the 322 organizations, 196 are treated as regional organizations because the organization title and/or charter specifically refer to a region, and 117 have a membership provision that restricts eligibility to countries in the region. 15. The dataset excludes organizations that are completely subsidiary to another IGO. Some eligibility restrictions are relaxed with a provision for a secondary track for accession, whereby a non-regional applicant or state not recognized as a member of the UN could become a member through a higher review process that often requires a two-third vote approval, while states in the main category of accession countries face an easier accession process.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 59

political system. Neither do IGO rules specify alliances as a precondition of membership.16 The practice of favoring fellow democracies and allies occurs almost entirely in the shadow of membership rules as informal discrimination to favor these states. Negotiable terms. The more flexible forms of membership restriction arise in the requirements for members to negotiate terms with an applicant for their accession. All but four of the 56 negotiable IGOs include economic topics. Negotiations at the time of accession specify the level of commitments in ways that allow for variation across members and within specific issues. For example, the WTO charter simply notes that “any State or separate customs territory possessing full autonomy in the conduct of its external commercial relations and of the other matters provided for in this Agreement and the Multilateral Trade Agreements may accede to this Agreement, on terms to be agreed between it and the WTO.” Accession negotiations have become lengthy sessions of bargaining over tariff schedules and periods to phase-in policy changes. A country such as China was led to make changes that exceed current member commitments in some areas while falling short of member commitments in other areas.17 In contrast, environmental topics are less likely to refer to negotiable terms of entry for accession—only five of the 59 IGOs that address topics related to the environment include negotiable accession terms. The original bargaining at IGO formation and for revision of the agreement is certainly negotiable, but the effort to include wide participation has inhibited use of negotiation at the time of accession. Member approval. A requirement for approval by members forms a general way to screen applicants without the need to specify criteria for admission. For example, the OECD Convention in Article 16 states: “The Council may decide to invite any Government prepared to assume the obligations of membership to accede to this Convention. Such decisions shall be unanimous.” This accession rule is quite open in its terms of eligibility, but restrictive to require invitation and unanimous approval by members. Thirty-eight percent of the IGOs require member approval by a vote, with some calling for a majority, supermajority, or consensus. By issue area, security IGOs are significantly 16. The exceptions are a subset of the security IGOs that form a security alliance, such as NATO and the Warsaw Treaty Organization. 17. See Yamaoka (2013) for details on WTO-plus and WTO-minus commitments in the Chinese accession protocol.

60

chapter 2

300

Number of IGOs

250 200 150 100 50 0

Selective

Eligibility

Member approval

Negotiable

Compliance review

figure 2.4. Restrictions on participation mandate: The figure shows the number of IGOs by the type of membership restriction present in the charter. Note that any single IGO may include multiple types of restrictions.

more likely than others to require member approval, with 54 percent having this requirement. But there is not as significant a difference across the other issue areas, where about a third of the organizations include a requirement for member approval. Thirty-eight percent of economic IGOs relative to 30 percent of environmental IGOs call for member approval.18 IGOs vary in the threshold of member approval specified in the charter document. Among the 122 IGOs with a member approval requirement, only 22 of them accept a simple majority vote. More common is a stringent approval requirement, which calls for a supermajority or unanimous consent of existing member states to admit new members (100 IGOs). These stringent approval IGOs have the same average membership as the full sample and are distributed across the issue areas. They form about one-third of regional IGOs (57 of 196) and also form about one-third of open IGOs that do not have any restrictions on eligibility (29 of 88). Selective IGOs. Restrictions on eligibility, negotiable terms, and member approval together differentiate selective IGOs from inclusive organizations. Figure 2.4 shows in the first column the number of IGOs that are selective 18. Only eight of the 61 IGOs with environmental mandate include an approval requirement. Thirty-five of the 101 social topic IGOs include an approval requirement. A comparison of means test shows that the only significant difference from population average is in security IGOs.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 61

on at least one of the three dimensions of restricted participation. With 282 of the 322 IGOs including some form of participation restriction, selective IGOs are the overwhelming majority (88 percent of the sample). In a hypothesis test for the IGOs in each issue area, only economic IGOs show a significant difference in their high use of selection relative to other IGOs.19 But this does not mean that the social and environmental issues are largely addressed within inclusive organizations—many selective organizations include a mandate for addressing social and environmental issues. The figure shows the distribution across each of the three types of participation restrictions examined separately, showing the most frequent use of eligibility restrictions, followed by member approval and negotiation. This pattern is similar across issue areas, with the one exception that security IGOs are more likely to rely on member approval than eligibility. For comparative perspective, the final column of Figure 2.4 shows that the use of compliance review as a form of conditionality to limit entry is much less frequent than the use of participation restrictions.

2.3.3

Conditionality Terms

Conditionality arises through two processes. First, states can specify the review of policies as part of the accession process. As distinct from negotiating the terms of entry, precision about conditionality represents a formal process to assess whether the country has changed its policies to comply with the treaty when it deposits the ratification instrument to become a member. Second, suspension and expulsion in the case of noncompliance forms the flip side of requiring compliance at the time of entry. The resolution of disputes over individual policy questions may go forward without implications for membership status, but conditionality terms set the expectation that some level of noncompliance could revoke membership. Meritocratic IGOs use conditionality to ensure that all states pass the required benchmarks for compliance. The Treaty on European Union signed in Maastricht in 1992 provides for accession in Article O: Any European State may apply to become a Member of the Union. It shall address its application to the Council, which shall act unanimously after 19. Among 256 economic IGOs, 230 are selective (90 percent) while among the 61 environment IGOs, 51 are selective (84 percent); among the 101 social IGOs, 88 are selective (87 percent); and, finally, among the 35 security IGOs, 31 are selective (89 percent).

62

chapter 2

consulting the Commission and after receiving the assent of the European Parliament, which shall act by an absolute majority of its component members. The conditions of admission and the adjustments to the Treaties on which the Union is founded which such admission entails shall be the subject of an agreement between the Member States and the applicant State. This agreement shall be submitted for ratification by all the contracting States in accordance with their respective constitutional requirements. The conditions of admission and adjustments to the treaties are separately acknowledged because the former include negotiable terms not formally laid out in the treaties while the latter affirms that policies are in full accordance with the treaties at time of admission. The treaty does not include provisions for expulsion from membership for noncompliance. The IAEA implies the threat of suspension in Article XIX as a form of conditionality: A member which has persistently violated the provisions of this Statute or of any agreement entered into by it pursuant to this Statute may be suspended from the exercise of the privileges and rights of membership by the General Conference acting by a two-thirds majority of the members present and voting upon recommendation by the Board of Governors. The models of IGOs represent an ideal type, and individual cases may fall between categories, with an IGO design that contains elements of multiple policy restrictions. For example, the EU holds weighted voting for hierarchy and imposes meritocratic conditionality as well as more discretionary member approval. Accession review process. Few IGOs specify the need for a compliance review process as a condition for accession at the charter level. The European Union treaty refers to the need for an agreement between members and the applicant detailing the “adjustments to the treaty.” Some will include reference to an assessment of intentions, but fall short of requiring a review of policies to confirm adherence to the treaty. In this category, one finds IGOs like the IAEA, which states: “In recommending and approving a State for membership, the Board of Governors and the General Conference shall determine that the State is able and willing to carry out the obligations of membership in the Agency, giving due consideration to its ability and willingness to act in accordance with the purposes and principles of the Charter of the United Nations.” But most IGOs simply require that the applicant deposit an instrument of ratification, and do not mandate a compliance review.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 63

The lack of conditionality within the terms of IGO charters does not mean that states disregard compliance. Members may determine through informal assessment that a state is far from compliance and vote against its membership, or they may use the negotiation of terms for entry as the means to impose conditionality. For example, the WTO establishes a working party of member states that reviews a document prepared by accession candidates on their trade policies and drafts an accession protocol that is brought to the full membership for approval. These procedures for reviewing applicants for admission typically lie outside of charter provisions. Including the formal requirement for review of policy conformity for membership represents a higher threshold of conditionality. Expulsion terms. We also observe conditionality through provisions for the suspension of membership in cases of noncompliance. For example, the West African Monetary Union charter in Article 4 declares that, “subject to automatic debarment from the Union, the States signatories shall undertake to honour the provisions of this Treaty and of enactments adopted for its implementation. . . . The Conference of Heads of State of the Union, by unanimous decision of the Heads of State of the other members of the Union, shall confirm the withdrawal therefrom of a State which has not honoured the above commitments.” Thirty-one percent of the IGOs (100 of 322) include a suspension clause; among these, 66 refer to nonpayment of dues as grounds for suspension, and 42 refer to noncompliance with the rules and norms of the organization. For suspension, states may lose the right to attend meetings and receive benefits of the organization on a temporary basis. In a more extreme provision, 62 organizations call for expulsion from the organization with terms to remove membership status. This often follows a period of suspension and requires a vote of the membership. The decision to include such rules is a sensitive matter. Emmons (2022) looks at the OAS deliberation over the amendment of institution rules in 1992 to add a Washington Protocol that included terms for suspension of governments that reversed on the organization commitment to democratic representation. States that had experienced external intervention feared that IGO terms for suspension could cover for another round of coercive moves by regional powers against their sovereignty, which led them to favor a high threshold for adoption, and led many to reject the protocol with a reservation. This example illustrates the problem with using the charter as a contract to enforce compliance. Expulsion provisions could safeguard against

64

chapter 2

backsliding, and would be most valuable amongst a heterogenous group of countries. But the diverse preferences and potential for disagreement in such a group will also make states reluctant to establish such terms. Calls for expulsion often meet the reality that the rules set too high a bar. The Russian invasion of Ukraine prompted many governments to urge rejection of its membership in international organizations. The Council of Europe, which has provisions for suspension and termination of membership for actions that violate its core commitment to rule of law and human rights, voted 15 March 2022 to expel Russia.20 In other organizations, this was more difficult. When the United States and its allies sought to suspend Russian membership in the International Criminal Police Commission, the secretariat office responded that there were no rules for ending membership and such decisions would require agreement of 195 members.21 As part of economic sanctions, governments sought to expel Russia from multilateral economic organizations, which also met with difficulty. The U.S. House approved legislation urging the USTR to “consider further steps with the view to suspend the Russian Federation’s participation in the WTO” and the European Commission president declared that “we will also work to suspend Russia’s membership rights in leading multilateral financial institutions, including the International Monetary Fund and the World Bank.”22 Since the WTO is among those organizations lacking any provision for expelling members, WTO experts supportive of expelling Russia explained that the path forward would be for two thirds of members to vote for an amendment to the WTO rules, creating a provision for expelling Russia.23 It was much easier for states to undertake actions to halt the WTO accession negotiations of Belarus for its support of the Russian invasion than to expel Russia. The IMF charter provides for expulsion of a member, but limits this to follow after members vote for a period of suspension and find “the member persists in its failure to 20. Council of Europe Newsroom, “The Russian Federation is excluded from the Council of Europe,” https://www.coe.int/en/web/portal/-/the-russian-federation-is-excluded-from-the -council-of-europe. Accessed 19 March 2022. 21. Darryl Coote, “Intel Alliance Asks Interpol to Suspend Russia as Kyiv Wants Moscow Isolated,” UPI, March 7, 2022. 22. House Committee on Ways and Means GOP, “Brady, Neal Introduce Bill to Further Punish Putin: Legislation Suspends Normal Trade Relations with Russia and Belarus,” March 17, 2022; European Commission, “Statement by President von der Leyen on the Fourth Package of Restrictive Measures against Russia,” 11 March 2022. 23. Bacchus, James, “Boot Russia from the WTO,” The Wall Street Journal, 28 February 2022.

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 65 200

Contains expulsion/suspension clause No expulsion/suspension clause

Number of IGOs

150

100

50

0

Economic

Security

Environmental

Social

figure 2.5. Membership suspension and expulsion: The figure shows the number of IGOs that include provisions in the charter for the suspension or removal of membership status. IGOs are grouped by issue area with the potential for an IGO to appear in more than one category.

fulfill any of its obligations under this Agreement.” Membership termination requires approval by an 88 percent voting share of the Board of Governors. That threshold requires support not only of the United States, but also China. An EU official acknowledged the challenge of expelling Russia from the IMF, noting, “The consensus legal reading is that it cannot be done. Even for members guilty of genocide it was not done, because it would require proof of a legal breach of the IMF agreement, so it [is] very unlikely.”24 One might expect some variation by the nature of the issue. Koremenos (2005) finds that states set shorter terms to treaties in some issue areas as a way to ex ante limit their commitment. A similar logic could extend to provisions for exit and suspension, but that is not what I find. Figure 2.5 shows by issue area the use of conditional membership through expulsion and suspension provisions in the charter, with comparison of the relative frequency of such provisions within the IGOs of each issue area. In none of the issue areas does the use of expulsion/suspension reveal a significantly higher or lower prevalence in comparison with the full population. 24. Reuters, “Exclusive: EU May Curb Russia’s Rights in IMF,” 4 March 2022.

66

chapter 2

Exit from IGOs is relatively rare, and most of these cases occur as voluntary exit rather than expulsion. Seventy-seven percent of the IGOs include a provision for exit from membership. Nevertheless, even these organizations average only four exit cases over the lifespan of the organization’s existence. Hierarchical decision-making. Hierarchical IGOs establish two or more classes of membership to uphold conditionality while having differential expectations of members. The provisions in the charter on membership are not restrictive on eligibility per se, although they may include more approval steps than the truly universal IGOs. Rather, in the allocation of decision-making authority within the IGO, the charter establishes differential influence. Many of the international financial institutions, including the IMF, World Bank, and Bank for International Settlements, adopt weighted voting based on the share of the funds contributed. Some commodity organizations allow for membership by both exporters and importers, but establish voting rules that distinguish between them. For example, the International Cocoa Organization fixes the total weight of voting rights available to exporters as a group relative to importers, and within these groups allocates weighted voting by share of exports/imports. The United Nations provides for a fixed hierarchy in the Security Council, which arises from the distinction between permanent members with a veto right and nonpermanent members elected from the full membership. As seen in Figure 2.6, economic IGOs are the most likely to implement weighted voting, and development banks stand out within this population. When an issue area confronts multiple institutions, weighted voting offers flexibility to accommodate asymmetric and changing power relationships (Lipscy, 2017). But the weights achieve control for contributors rather than serving as a tax on utilization. The differentiated membership supports broader participation by allowing those who pay more funds to exercise influence while lightening the burden on other states that will be drawing on the services of the IGO most heavily. Having examined each component of charter design that restricts participation and imposes conditions, we can revisit the four conceptual models of IGO accession presented in Figure 2.1 at the beginning of the chapter. The pie chart in Figure 2.7 shows the breakdown when classifying each of the 322 IGOs into one model based on their charter provisions for membership. The universal IGOs, which are in principle open to any state, do not restrict eligibility on region, commodity production, or other criteria, hold a

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 67 200 Weighted vote Non-weighted vote

Number of IGOs

150

100

50

0 Economic

Social

Environmental

Security

figure 2.6. Weighted voting: The figure shows the number of IGOs with weighted voting by issue area.

low threshold of member approval, and do not impose conditions to screen compliance or differentiate among members with weighted voting. In the hierarchical model, weighted voting differentiates among members, and there may also be a member approval requirement. Meritocratic IGOs are highly selective as they make membership conditional on inspecting that states implement rules. The club model is also selective through limiting eligibility and/or requiring member approval but does not require review of policies as a condition of membership. The dominant design choice for accession in IGO charters is selective participation and vague conditionality, with nearly 80 percent of the IGOs falling into the category of the club model accession process. Accession also includes informal practices. The OECD, WTO, and Asian Development Bank (ADB) do not refer to compliance review in their charter, but lay out a rigorous accession review process with detailed information available on their website. Yet, expanding the compliance review to include practices mentioned on the IGO’s websites, and shifting these IGO’s classification from club to meritocratic, would expand the share of meritocratic IGOs from seven to only 19, and remain a minority practice. While other IGOs may conduct screening, they choose not to make the compliance review explicit as a condition for entry either through the charter or website. Such choices let members raise or lower the entry conditions as they see fit. The charter

68

chapter 2 Meritocratic 2% Hierarchical 7%

Universal 12%

Club 79%

figure 2.7. Accession patterns: The figure shows the share of IGOs according to their category of accession process.

design reveals state preferences for a club model that provides wide discretion to determine which states qualify to join.

2.4 Toward a Broader Understanding of Accession This survey of the design of membership provisions contained in IGO charters shows two clear patterns. First, universal IGOs are a small share of the IGO population. Most IGOs restrict membership. Second, when limiting entry, they rely more on participation rules than conditionality. Neither the formal review of policy compliance nor the threat of expulsion for noncompliance are commonly used. The exclusion mechanisms uphold discretion for members to change their criteria for screening members. Member approval and negotiable terms leave the widest flexibility for raising and lowering the entry bar. Yet even exclusive eligibility rules appear to be vague in their categories rather than rigid performance criteria or fixed categories of states. The most limiting category of regional integration itself has been flexible through fluid definitions of region and allowance for participation by states outside of the region for some regional organizations. The IGOs that address environmental topics that come closest to public goods are more likely to have inclusive participation rules and are less likely to impose conditionality. Yet it is not true that they are all large in membership

f l e x i b i l i t y b y d e s i g n : r u l e s f o r a c c e s s i o n 69

and uphold a universal accession process. Indeed, the average size of environmental IGOs is smaller than that of others. Economic IGOs show a strong pattern for restrictive membership and stand out for use of exclusive eligibility, negotiable terms, and weighted voting. On dimensions of member approval and use of expulsion/suspension provisions, the economic IGOs resemble those in other issue areas. These could be limiting factors that facilitate discriminatory practices for economic IGOs and reduce the room for discrimination in environmental IGOs. But the differences would be a matter of degree. Broadly speaking, the analysis of charters reveals that deference to states to select members occurs across issue areas. There are several caveats with which we must view the analysis of IGO charters. These documents are open to change. Few have revised their rules on membership, but this is an area for future research into the evolution of membership rules. More challenging is the role of administrative procedures and informal practices that shape accession processes outside of the strictures of the founding documents. States have chosen to leave this area of IGO rules vague in their constituting documents. Instead of setting ex ante conditions for entry, they establish decision rules at a lower level and fill in gaps with informal practices.

2.5 Conclusion This chapter has explained the prevalence of club-style design of international organizations. As states seek to maximize both the value of association with particular states and the gains of cooperation on important policy issues, they have prioritized flexibility over membership decisions. The evidence from analysis of IGO charters strongly supports the hypothesis for discretionary rules. This lays the baseline institutional context for geopolitical discrimination. Because IGO membership rules are more often discretionary than rigid, members can inject alternative criteria into their decisions to select members. If the OECD mandated democratic governance, or if the organization engaged in rigorous performance review of member compliance with all its codes and conventions, states such as Turkey and Mexico may have had more difficulty in becoming members. If the WTO were truly universal, the long negotiations with China or exclusion of Iran over foreign policy objectives unrelated to trade would not have held up the accession of those countries. Instead, flexibility in the exclusion mechanism applied within IGO charters has allowed unexpected entrants and the denial of unwanted applicants.

70

chapter 2

The advantage of the club model for selective but discretionary design of accession lies in its flexibility. This both allows members to discriminate and applicants to differentiate themselves as belonging to a special group. Yet such discretion comes at a cost as states eschew forming a rigid contract that defines terms for entry and expulsion as leverage to achieve higher levels of compliance. States have been willing to accept this trade-off across a large number of IGOs. Although environmental IGOs are more likely to adopt a universal scope and economic IGOs are more likely to adopt selective entry, the use of club design holds across all issue areas. Moreover, the mechanism of selection is voting approval. Compliance reviews are rare, and, when taken, exist as an informal procedure adopted outside of the formal charter rules for membership. The extensive reviews of policy changes to impose conditionality on EU applicants for the Eastern enlargement represent an unusually rigorous accession process, and such meritocratic evaluation is far from the norm. Yet neither do IGOs allow any applicant to join. In the gray zone between universality and compliance evaluation lies discrimination—choosing to cooperate with those states that form a community. Next we turn to examine the unwritten criteria for membership and explain how and why security interests impact global governance through the politics of IGO membership.

3 Membership Patterns in Economic Institutions Christina L. Davis and Tyler Pratt1

during the formation and enlargement of international organizations, membership depends on states’ specific attributes relevant to the regulated issue as wellastheirbroaderpoliticalrelations.Themainthemeofthebookengagesthe idea that geopolitical discrimination occurs when states selectively allow those who share their foreign policy interests to enter international organizations. This chapter measures the timing and weight allocated to geopolitics in the membership choices for multilateral economic organizations. We demonstrate that geopolitical discrimination has been a prevalent practice across a broad swathe of economic organizations for over 60 years. In this role, shared security interests form a foundation for economic cooperation. This builds upon a wide range of research into the interaction between states’ economic and security interests (e.g. Gowa and Mansfield, 1993; Mansfield and Bronson, 1997; Mansfield and Pollins, 2003; Lake, 2009). During the Cold War, states allocated aid according to the military importance of recipients, and new strategic needs shape development policies today (e.g., Meernik, Krueger, and Poe, 1998; Bearce and Tirone, 2010; Boutton and Carter, 2014; Bermeo, 2018). Like aid, trade flows tend to follow the flag (e.g., Pollins, 1989; Keshk, Reuveny and Pollins, 2004; Pandya, 2016). The income gains from trade spillover into military power, motivating states to trade with allies 1. This chapter is based on the article Davis, Christina L. and Tyler Pratt. 2021. “The Forces of Attraction: How Security Interests Shape Membership in Economic Institutions.” The Review of International Organizations 16:903–929. 71

72

chapter 3

(Gowa, 1989; Gowa and Mansfield, 1993; Gowa, 1994; Mansfield and Bronson, 1997; Long and Leeds, 2006). Other sources of influence can skew trade toward states with good relations (Berger et al., 2013; Fuchs and Klann, 2013). We look beyond bilateral exchanges of trade and aid during the Cold War to examine how security linkages shape multilateral economic governance. The use of geopolitical discrimination when states join multilateral economic institutions resolves a dilemma for states. On the one hand, the principle of nondiscrimination and practice of collective decision-making render multilateral institutions inefficient as tools of economic statecraft. It is much easier for states to manipulate bilateral aid and trade policies where national governments wield control. On the other hand, multilateral institutions govern a large amount of economic activity, and states have incentives to exploit these gains for geopolitical advantage. Discrimination over membership allows states to steer benefits under the guise of multilateralism. Through this practice, powerful states build their geopolitical coalitions and weaker states minimize the reforms needed to gain entry. Section 3.1 connects the theory about geopolitical discrimination to the empirical test on multilateral economic organizations. We specify observable implications to assess the theory’s predictions for patterns of membership. In section 3.2, we introduce our data on economic IGOs and present the results of ourempiricalanalysisofmembership.Thechapterexaminesasampleofpotential memberships in 231 multilateral economic organizations for cases when existing members have discretion over new members. We develop a relational measure of geopolitical alignment based on alliance ties and UN voting similarity between potential members and other states in the organization. Trade flows between new states and other members serve as a proxy for the functional theory that economic interdependence explains demand for institutionalized cooperation on economic issues. We find that measures of geopolitical alignment between countries are strongly predictive of membership even when controlling for trade ties. Where earlier work focused only on enlargement, our evidence reveals that geopolitical discrimination occurs during the formation and enlargement of economic organizations. After confirming that both geopolitical alignment and economic interdependence drive membership, we use a finite mixture model to estimate the relative weight accorded to security versus trade ties. This model specification allows us to estimate how much states careaboutsecuritytieswhenmakingmembershipdecisions,whilerecognizing that economic interests also play an important role. We find that geopolitical alignment motivates 44 percent of the membership decisions in our sample of

m e m b e r s h i p patt e r n s i n e c o n o m i c i n s t i t u t i o n s 73

231 multilateral economic organizations over the period 1949–2014. Far from being the exception, security ties are a prominent factor that shapes who joins and remains in multilateral economic organizations.

3.1 Geopolitics and Economic Cooperation Economic interdependence generates potential mutual gains, and states have incentives for cooperation to harness these benefits. Functional theories emphasize the ability of institutions to overcome market failures that hinder cooperation (Keohane and Nye, 1977; Keohane, 1984; Farrell and Newman, 2015; Koremenos, 2016). The long-term gains from setting economic policies in cooperation are supported by commitment and enforcement mechanisms in multilateral economic organizations that raise the cost of unilateral policies that might otherwise tempt states to capture short-term gains. Eighty percent of IGOs with formal charters include economic cooperation within their mandate. Economic interests are a prime driver of membership in these organizations. As a baseline expectation, membership selection should privilege those with the economic profile conducive to cooperation in the issue area regulated by the institution. A performance evaluation would focus on the many available economic indicators. For example, development banks might review financial accounts to evaluate the need for a country to receive development assistance and its ability to repay creditors, or to assess the potential of a government to serve as a donor. Trade volume could indicate the potential value in accepting a state into a trade organization, alongside reviews of marketbased economic regulations. The specific economic criteria could be tailored to the mandate of each organization. Discretion over membership, however, allows states to politicize institutional outcomes and choose members based on noneconomic criteria. When IGOs become discriminatory clubs, membership reflects both the ability to contribute to the joint project and an intrinsic value to the group. From this perspective, the appeal of joining an IGO depends on the composition of its members and not just mutual interests on a narrow issue. While states could choose among many criteria when discriminating, the geopolitical discrimination hypothesis posits that security interests form a common source of discrimination in IGO membership. Both security externalities and security linkages undergird the logic. The security externality mechanism focuses on the inherent connection between economic and military

74

chapter 3

power. Efficiency gains from economic cooperation can be channeled into military power (Gowa, 1989; Gowa and Mansfield, 1993; Gowa, 1994). States recognize these spillovers and respond strategically, steering the benefits of multilateral economic cooperation toward allies and away from adversaries. Whereas the security externality arises as an indirect byproduct of economic exchange, a security linkage represents a joint decision that brings economic and security interests together. In this strategy, membership becomes a bargaining chip for states to expand their geopolitical coalitions. Different preferences for security and economic cooperation create the opportunity for a mutually profitable exchange (Tollison and Willett, 1979; Davis, 2004; Poast, 2013). Some states may be more motivated by the prospective economic gains from membership, while others seek enhanced leverage on security matters. The institutional context strengthens the linkage by providing an exclusion mechanism that restricts cooperation to the subset of members and increases the credibility of promised benefits to recipients. We recognize that state preferences may arise from both functional demand for gains from regime cooperation and geostrategic goals. States must balance the desire to exploit selection of members for geopolitical gain with the functional trade-offs. The key task is differentiating their relative impact on outcomes.

3.1.1 Testable Implications for Membership Patterns The empirical analysis examines whether states are more likely to join international organizations when they share geopolitical alignment with current members. This examines the second hypothesis presented in chapter 1: States with shared geopolitical alignment form organizations together and are more likely to join the same organizations. Our research design uses observational data about state behavior, making it hard to identify causal effects. To reduce the risk of a spurious finding, we use multiple measures of geopolitical alignment, explore alternative model specifications, and exploit dynamic breaks in alignment to more closely identify the mechanism. We compare geopolitical alignment with economic interdependence as two forces that influence the demand for membership. Since geopolitical alignment and interests within the issue area overlap entirely in the area of security organizations, looking at economic organizations provides a better test. Our argument contends that security interests shape membership even

m e m b e r s h i p patt e r n s i n e c o n o m i c i n s t i t u t i o n s 75

in multilateral economic organizations, while the functional perspective would suggest a smaller role. We use shared alliances as our primary operationalization of geopolitical alignment, and we analyze similarity of alliance portfolios and UN voting patterns as two alternative measures. Trade ties between states control for the underlying interdependence that contributes to functional demand for cooperation on economic policies. We also examine conditions that influence the importance of security interests. First, we distinguish between geopolitical alignment with all members of international organizations versus the most powerful member states. In principle, any existing member state can link accession of potential members to shared security interests. Below, we measure a state’s geopolitical ties with all IGO members to assess how comprehensively the state shares foreign policy interests with existing members. Then we separately measure a state’s ties with powerful states, who are the most likely to engage in linkage strategies as they cement relationships with strategic partners (Schneider and Urpelainen, 2012). Our relational measures of alignment differ from the monadic approach of current studies. For example, Kaoutzanis, Poast, and Urpelainen (2016) focus on the regime type of new entrants and Donno, Metzger, and Russett (2015) examine their conflict history. Second, we analyze whether the role of geopolitical alignment differs over the lifespan of an IGO. Establishing an IGO raises transaction costs, as states negotiate the IGO charter and set up a headquarters and financial base. While states may use security linkages to overcome the initial cooperation challenges, they could focus strictly on policy when establishing the rules of the game and only later politicize membership. Prior work on screening members has been limited to enlargement (Kydd, 2001; Schneider, 2009; Stone, 2011; Donno, Metzger, and Russett, 2015). Through separate analysis of each stage, we test how geopolitics shapes membership at founding and enlargement.

3.2 Empirical Analysis of IGO Membership Patterns Our empirical tests have two primary objectives. First, we estimate the effect of geopolitical alignment on the probability of IGO membership in a series of regression models, controlling for functional economic interests. Second, we estimate the relative weight given to geopolitical and functional considerations. We use a finite mixture model to assess which observations are more

76

chapter 3

consistent with geopolitics or economic interdependence, and to show the conditions under which states privilege one over the other.2

3.2.1

Data on Membership in Multilateral Economic Organizations

We analyze a sample of 231 economic IGOs for the period from 1949 to 2014.3 Starting with the IGO membership data introduced in chapter 2, we subset to those IGOs with an economic focus.4 We conduct our analysis at the level of the state-IGO-year (Poast and Urpelainen, 2013; Donno, Metzger, and Russett, 2015). This unit of observation reflects the decision process we seek to model: when a state elects to form or join an IGO, it considers a specific group of other states. Because we are interested in how states exploit discretion over membership, we limit our sample to “potential but not automatic” IGO memberships. This entails two exclusions from the set of all possible state-IGO-year pairings in the period 1949–2014.5 First, we exclude universal IGOs, where existing members have less discretion to select new entrants.6 This removes organizations where membership is functionally automatic for states that seek entry, like the UPU or International Maritime Organization (IMO). Second, we exclude state-IGO pairings for regional organizations where the state resides outside the regional scope of the IGO.7 As a result, the sample does not include memberships with little chance of being instantiated 2. See the supplemental materials in Davis and Pratt (2021) for tables including robustness checks. 3. Economic topics are broadly construed to include aid, finance, trade, and more general management of resources, travel, and standards relevant for economic exchange. Our coding of organizations draws on the Yearbook of International Organizations’ description of the aims and subject of each organization along with information contained in the IGO charter documents. 4. The main dataset draws upon the Correlates of War (COW) International Organizations Dataset (Pevehouse, Nordstrom, and Warnke, 2004), with additional filtering based on coding of the organizations with a formal IGO charter document. Results are not contingent on our exclusion of IGOs without charter documents. 5. Our empirical results are robust to the inclusion of these observations. 6. Chapter 8 applies the theory to open eligibility organizations, including the universal IGOs. The analysis of this chapter has also been applied to the full sample of all IGOs; see appendix tables in Davis and Pratt (2021). 7. See chapter 7 for separate analysis of regional IGO membership, including non-regional states as potential entrants.

m e m b e r s h i p patt e r n s i n e c o n o m i c i n s t i t u t i o n s 77

(e.g., Swedish membership in the AU). This generates 570,695 state-IGO-year observations.8 The dependent variable, IGO Membershipijt , is a dichotomous measure of whether state i is a member of organization j in year t. IGO Membership is equal to 1 in 36.9% of observations. In the tests below, we also subset this sample to assess whether geopolitical alignment has different effects on joining as a founding member, joining later by accession, or exiting the organization. Our theory suggests security externalities and linkage will shape incentives for both founding members and those joining later, so our main sample pools these two paths to membership. Yet, to account for potential differences in these underlying decision processes and address alternative approaches in the literature, we also test our argument separately within each subset.9 Formal alliances are our primary measure of geopolitical alignment.10 The variable Average Alliancesijt measures the proportion of IGO j’s member states with which state i shares a formal alliance in a given year. In the sample, it ranges from 0 (49.7% of observations) to 1 (4.5% of observations), with a mean value of 0.15. Second, Lead State Allianceijt indicates whether state i shares an alliance with the leading economic power among member states of IGO j during year t, with economic power measured annually by gross domestic product (GDP). In our sample, states are allied with an IGO’s most powerful member state in 23% of state-IGO-years. We also analyze two alternative measures of geopolitical alignment. S-scores is a continuous measure of similarity across states’ entire portfolio of alliances; it reaches its maximum (1) when two states have identical alliance portfolios.11 UN Ideal Point Similarity is a continuous variable that increases as the UN voting records of two states converge (Bailey, Strezhnev, and Voeten, 2017). 8. IGOs enter the dataset in the year in which they are founded and continue until 2014 or the year that the organization ends. We include all states listed in the COW state system for which we have data on covariates. Covariate coverage primarily excludes small states (e.g., Grenada) or those for which data is unavailable (North Korea). We follow the COW coding for the start and end of IGO existence. 9. Donno, Metzger, and Russett (2015) focus their analysis of IGO accession on the enlargement phase, and Poast and Urpelainen (2013) demonstrate that the politics of forming new IGOs differs from that of joining existing IGOs. 10. Data on alliances, which include defense pacts and neutrality or nonaggression pacts, come from version 4.1 of the COW Formal Alliances dataset (Gibler, 2009). 11. S-scores are calculated using the COW formal alliance dataset.

78

chapter 3

This measure offers a broader perspective on the foreign policy orientation of states across a range of topics on the international agenda, and it has been widely used in the literature to measure geopolitical alignment (e.g., Bearce and Bondanella, 2007; Vreeland and Dreher, 2014). Each alternative measure is operationalized to create both an “average” and “lead state” variable. Trade with IGO members and trade with the IGO lead state measure the impact of shared economic ties.12 We include monadic variables for income (GDP per capita, logged), market size (GDP, logged), and trade openness (total trade / GDP).13 Conditioning on these economic variables addresses the possibility that economic flows and security interests are jointly determined. We control for additional variables that influence the demand to join IGOs and the willingness of members to grant entry. Polity scores capture the tendency of democratic states to join and form IGOs with higher frequency (Russett and Oneal, 2001; Poast and Urpelainen, 2015). We account for the screening out of conflict-prone states (Donno, Metzger, and Russett, 2015) by including a variable measuring the number of fatal interstate militarized disputes (MIDs) between state i and members of IGO j.14 To address potential diffusion effects, the variable Total IGO Membership is a count of members in each IGO, which could exert positive attraction for other states to enter. Since neighbors may exert stronger influence over states, we also include a variable for Members from Region, indicating the number of states residing in state i’s geographic region that are members of IGO j. Separate control variables measure shared colonial history as well as a state’s average geographic distance from IGO j’s member states.15 Finally, the design of the IGO influences its openness to new members. An indicator for Stringent Approval identifies organizations that require a supermajority or unanimous consent of existing member states to admit new 12. Bilateral trade data is from the IMF Direction of Trade dataset. The “trade with members” variable measures average (logged) volume of merchandise imports and exports between state i and each member of IGO j. The “trade with lead state” variable measures (logged) trade volume with the lead state. We add 1 before taking the log to ensure values of zero trade are not excluded due to the mathematical transformation. 13. We use the natural log of GDP and GDP per capita in constant 1967 U.S. dollars. Data through 2004 are from Goldstein, Rivers, and Tomz (2007); we use adjusted World Bank GDP estimates to fill in subsequent years. 14. MIDs data are from the dyadic version of the COW Militarized Interstate Disputes Dataset. 15. Data on geographic distance and colonial linkages are from CEPII.

m e m b e r s h i p patt e r n s i n e c o n o m i c i n s t i t u t i o n s 79

members, according to the founding charter.16 We include an indicator for regional organizations and follow Carter and Signorino (2010) in modeling time dependence with a cubic polynomial for t in all models. A Cold War indicator (1947–1991) adjusts for baseline differences in membership rates during the bipolar era.

3.2.2

Logistic Regression Analysis of Membership

We first use a logistic regression model to predict the dichotomous outcome variable, IGO membership. Independent variables are lagged by one year, and standard errors are clustered at the country level. We estimate the following model of IGO membership for state i in IGO j and year t: Pr(IGO membershipijt = 1) = logit −1 (α + β1 Alliancesijt−1 + β2 Xijt−1 ). The model predicts IGO membership using our primary operationalization of geopolitical alignment, formal alliances. All models include a set of control variables Xijt , which are measured at the level of the state-IGO-year (e.g., Trade and fatal MIDs with IGO Members); state-year (Trade Openness, GDP, GDP per capita, Polity); and IGO (Stringent Accession, Regional IGO). We begin using the Average Alliances measure of geopolitical alignment. Table 3.1 displays results for a reduced form specification (Model 1) and a full model with all control variables (Model 2). In these first specifications, we assess membership in the broadest sense, including a state’s entry into an IGO and each year of continued membership.17 The results support our primary hypothesis: as states share more alliances with an IGO’s member states, they are significantly more likely to join the organization. In the full model, a one standard deviation increase in the Average Alliances measure increases the probability of membership, on average, from 35% to 41%. The relationship holds when controlling for functional economic interests (measured by trade with IGO members), which also has a positive and significant association with IGO membership.18 16. We also include IGOs that require potential members to receive approval from a specific committee. Approximately a third of the IGOs in our sample (75) have stringent approval procedures. 17. This is consistent with Stone (2011), who theorizes participation in IGOs as an ongoing process of decisions to enter and continue cooperation. 18. States’ alliance and trade ties are positively correlated in our sample (0.18), consistent with existing work demonstrating that allies are more likely to trade with each other (Gowa and

570,695 231 164

570,695 231 164

1.949∗∗∗ (0.171) 0.296∗∗∗ (0.033) 0.011∗ (0.006) 0.053 (0.062) −0.245∗∗∗ (0.043) −0.014∗∗∗ (0.005) 0.069 (0.049) 0.325∗∗∗ (0.102) −0.038 (0.039)

1.321∗∗∗ (0.113) 0.196∗∗∗ (0.010) 0.001 (0.004) −0.098∗∗∗ (0.018)

364,822 231 164

1.123∗∗∗ (0.050) 0.120∗∗∗ (0.006) 0.029∗∗∗ (0.002) 0.025∗∗ (0.011) −0.159∗∗∗ (0.015) −0.019∗∗∗ (0.006) −0.105∗∗∗ (0.029) 0.109∗∗∗ (0.037) −0.704∗∗∗ (0.062)

(3) Entry

10,590 193 164

1.296∗∗∗ (0.145) 0.168∗∗∗ (0.017) 0.019∗∗∗ (0.004) −0.060∗∗ (0.030) −0.225∗∗∗ (0.034) −0.036 (0.036) −0.514∗∗∗ (0.048) 0.121∗ (0.069) −0.340∗∗∗ (0.098)

(4) Formation

354,232 231 164

1.038∗∗∗ (0.083) 0.107∗∗∗ (0.009) 0.033∗∗∗ (0.003) 0.034∗∗ (0.015) −0.141∗∗∗ (0.020) −0.019∗∗∗ (0.006) 0.066∗ (0.039) 0.041 (0.058) −0.815∗∗∗ (0.088)

(5) Enlargement

211,039 227 164

−0.300 (0.262) −0.002 (0.037) −0.014 (0.010) 0.030 (0.060) −0.189∗∗∗ (0.069) −0.005 (0.017) −0.240∗ (0.135) −0.817∗∗∗ (0.199) −1.066∗∗∗ (0.280)

(6) Exit

364,822 231 164

−2.164∗∗∗ (0.185) −0.707∗∗∗ (0.074)

3.104∗∗∗ (0.114) 0.082∗∗∗ (0.009) 0.040∗∗∗ (0.004) −0.485∗∗∗ (0.106) 0.401∗∗∗ (0.101) −0.017∗∗∗ (0.006)

(7) State-IGO FE

570,695 231 164

0.335∗∗∗ (0.027) 0.012∗∗∗ (0.001) 0.002∗∗∗ (0.001) 0.022∗∗∗ (0.006) −0.015∗∗∗ (0.005) −0.0005∗∗ (0.0002) 0.012∗∗ (0.005) 0.040∗∗∗ (0.012)

(8) Diff-in-diff

table 3.1. Effect of alliances on IGO membership: Results of logit models estimating the effect of alliances on membership in economic IGOs. Coefficient estimates are displayed with robust standard errors in parentheses. Models 2–6 include the following controls (not shown): Fatal MIDs with Members, Members from Region, State-IGO Same Region, IGO Membership Size, Total State Memberships, Former Colony, Common Colonial History, and a time polynomial. Models 3, 5, 6, and 7 utilize rare events logit. Statistical significance is denoted by: ∗ p